"The Dinosaur with Good Proportions: The Origin Story of the P-47 Thunderbolt"

In the spring of 1940, Alexander Kartveli boarded a train at Washington's Union Station and headed back to New York with nothing to show for the trip. The Army Air Corps had recently evaluated his XP-44 Rocket design at Wright Field in Dayton and found it wanting, and the Washington conference that followed had made the conclusion official: the XP-44 concept was a dead end. Reports coming out of Europe painted a picture of air combat that made nearly every American fighter on the drawing boards look inadequate. The Germans were flying Messerschmitt Bf 109s that climbed faster and turned harder than the P-40, and a newer fighter called the Focke-Wulf Fw 190 was in prototype testing and would enter Luftwaffe service in 1941 to make things considerably worse. Kartveli's company currently had nothing on offer that could answer either of them.

He had been here before. His entire career since arriving in the United States a decade earlier had been a series of designs that were nearly good enough, almost competitive, consistently promising but never quite decisive. The P-35 had been an elegant aircraft but outclassed almost immediately by European developments. The AP-4 demonstrator had impressed the Army with its high-altitude performance, but the contract had still gone to Curtiss and its P-40. Now the XP-44 was dead before it flew. By the time the train pulled out of Union Station, according to the account that has passed through Republic Aviation's institutional history and multiple museum sources, Kartveli had a pencil in his hand and a pad of paper on his knee. The aircraft taking shape in those sketches would eventually be called the Thunderbolt. More of it would be built during the war years than any other American fighter. The story of how it got there is, at its core, the story of two immigrants from the same Russian imperial city who built something extraordinary together and then couldn't stay out of each other's way long enough to see it finished.

Alexander Nikolayevich Prokofiev de Seversky was born on June 7, 1894, in Tiflis in the Russian Empire, the city that would later become the capital of the independent Republic of Georgia. His father was one of the first civilian Russian aircraft owners, and young Alexander grew up with aviation not as an abstraction but as a practical part of his world. He graduated from the Imperial Russian Naval Academy in 1914 with an engineering degree, received his commission in the Imperial Russian Naval Air Service, and shipped out to the Baltic Fleet as a combat pilot in the summer of 1915. On his very first mission, attacking a German destroyer in the Gulf of Riga, his aircraft was shot down by antiaircraft fire before he could release his bombs. He was flying a Franco-British Aviation Company FBA flying boat, a wooden-hulled biplane seaplane powered by a 130-horsepower Clerget rotary engine driving a backward-facing pusher propeller, capable of barely exceeding 60 miles per hour. The bombs exploded in the crash. His observer was killed. Seversky lost his right leg below the knee.

Russian naval regulations removed him from flight status on the spot, and Seversky spent his convalescence period arguing with every officer who would listen that an artificial limb should not ground a pilot who could demonstrably still fly. He was making that argument in writing, in person, and with increasing frustration when a fortuitous coincidence intervened. Tsar Nicholas II, who happened to be inspecting the naval air station, encountered Seversky in the middle of an unauthorized aerial demonstration, having climbed into an aircraft and taken off without official clearance to prove his point. The Tsar's reaction to discovering a legless pilot performing spirited aerobatics over his inspection tour was, after the inevitable arrest, to intervene on Seversky's behalf. In July 1916, Seversky was returned to combat status, now flying the Grigorovich M-9, a Russian-built biplane flying boat that had replaced the FBA in Baltic Fleet service. He shot down his first enemy aircraft three days after returning to duty.

What followed was one of the more remarkable combat careers in Russian naval aviation. Seversky flew 57 combat missions and claimed 13 aerial victories. Researchers working from Imperial Russian records cross-referenced against German loss documentation have confirmed six of those victories. The discrepancy is genuine and worth understanding rather than simply noting. The Imperial Russian Air Service had no standardized aerial victory confirmation system of the kind that the British Royal Flying Corps and the German Air Service developed during the same war. Much of the fighting over the Gulf of Riga took place above open water, where a downed aircraft left no wreckage for ground forces to verify, and the confirmation standards that did exist were inconsistent from unit to unit and commander to commander. Seversky left Russia in 1918, before the revolution destroyed whatever central records existed, and the accounts he brought with him to America reflected his own claims rather than a reconciled official tally.

The honest accounting is that he claimed 13 and that six can be confirmed against surviving records. Those six confirmed kills place him at the top of the Russian naval aviation list specifically, since the higher-scoring Russian aces of the war, men like Vasili Yanchenko with 16 victories and Pavel Argeyev with 15, were Army aviation pilots rather than naval aviators. Among pilots who flew specifically for the Imperial Russian Navy, six appears to be the highest confirmed total in surviving records, which would make Seversky the leading naval ace of the conflict on documented kills alone, though specialists note that the incompleteness of Russian WWI records makes any such ranking provisional rather than definitive. If his claimed total of 13 is accepted, he would rank as the third-highest-scoring Russian aviator of the entire war across all branches. The true number is not known and cannot be recovered. What is not in dispute is that the Tsar personally presented him with a Gold Sword for valor, that he received the Order of St. George along with five other decorations, and that he commanded the 2nd Naval Fighter Detachment from February 1917 until a horse-drawn wagon, of all things, ran over his remaining good leg in a street accident and put him back in the hospital.

The revolution of 1917 ended that world. In March 1918, Seversky arrived in the United States as part of a Russian naval purchasing mission and, like many Russian emigres of his generation, found no compelling reason to return to a country that no longer existed in the form he had served. He offered his services to the War Department, and Major General William Kenly, Director of the Division of Military Aeronautics and the head of United States Army aviation in 1918, appointed him as a civilian consulting engineer and test pilot assigned to the Buffalo District of aircraft production. That civilian appointment, not a military commission, was the mechanism that got an amputee into American cockpits: he was hired for his aeronautical expertise, not enlisted through a physical examination. He worked in that civilian capacity for several years, filing patents at a rate that suggested someone had removed sleep from his daily schedule, and later worked alongside Billy Mitchell. He became an American citizen in 1927, and the following year was commissioned as a major in the Army Air Corps Reserve, a recognition of a decade of contributions to American aviation rather than a standard enlistment process. Over the following decade he accumulated more than 360 American patents covering split flaps, metal monocoque fuselage construction, fire-control systems for aircraft guns, retractable landing gear designs, specialized flight instruments, pontoon configurations, and an aerial refueling concept he had conceived while flying formation with another Russian pilot in the war, grabbing the trailing antenna wire and thinking about what a fuel tube in its place could accomplish. He set world speed records in an amphibious aircraft of his own design in 1933 and 1935. He was, by any honest measure, one of the most inventive aviation minds working in the United States.

When Seversky arrived in the United States in 1918, one of his first connections was to Brigadier General Billy Mitchell, the Army Air Service's most outspoken advocate for independent airpower. Seversky advised Mitchell during the famous 1921 bombing tests off the Virginia capes, in which Army bombers sank the captured German battleship Ostfriesland and several other vessels to demonstrate, in the most graphic terms available, that aircraft could destroy capital ships that the Navy had assured Congress were immune to aerial attack. Mitchell's court-martial four years later for publicly criticizing the Army and Navy's resistance to airpower development was a political confrontation that Seversky watched closely. His aerial refueling patent, which he filed in 1921 and which described a hose-and-grapple method for transferring fuel between aircraft in flight, was the technical foundation for the 1929 "Question Mark" endurance flight, in which a Fokker C-2A trimotor remained aloft continuously for just under seven days by refueling from Douglas C-1 transports using the Seversky system. Among the crew of the Question Mark were three officers who would shape American airpower in the Second World War: Carl Spaatz, Ira Eaker, and Elwood Quesada. The idea that had come to Seversky while grabbing the antenna wire of a fellow Russian pilot over the Baltic in 1916 had, thirteen years later, demonstrated endurance flight that changed how the American military thought about the range limitations of aircraft.

In February 1931, with backing from Wall Street investors, he founded the Seversky Aero Corporation in Farmingdale, New York, on Long Island. The company would specialize in high-speed, long-range pursuit aircraft. Seversky served as president and chief test pilot. The first thing he needed was talented engineering help, and in 1931 he made the hire that would ultimately define everything the company built.

Alexander Kartveli was born Aleksandre Kartvelishvili on September 9, 1896, also in Tiflis. He and Seversky were both sons of the same Russian imperial city, separated in age by two years, shaped by the same culture, and possessed of the same native language that neither would speak professionally again. Kartveli had pursued his technical education in France, studying at the Paris Aviation Higher College of Engineering and graduating in 1922, then working for Blériot Aéronautique before crossing the Atlantic in 1928 to join the Fokker American Company in Passaic, New Jersey. When Seversky found him, Kartveli was already one of the most technically rigorous aeronautical engineers working in the United States. Seversky hired him as chief engineer. Two Georgians from Tiflis, one a celebrated fighter pilot and promoter who filled every room he walked into, the other a methodical designer who trusted calculations over charm, set up shop together on Long Island and went to work.

Their collaboration produced aircraft of genuine quality from the beginning. The SEV-1XP, which Kartveli designed, outperformed the Curtiss P-36 Hawk in a 1936 Army Air Corps competition and led directly to the production contract for the Seversky P-35. It was the first single-seat fighter in United States Army Air Corps service to combine all-metal construction, retractable landing gear, and an enclosed cockpit in a single airframe, and it won the Bendix Trophy air race in 1937, 1938, and 1939 in civilian trim. Kartveli gave every aircraft he designed from the P-35 forward the same semi-elliptical wing planform, a signature he would carry from Long Island all the way through the Cold War. The P-35 was a genuinely modern fighter for 1936. It was also the aircraft that would, through a direct chain of evolution and one complete reinvention on a train ride north from Washington, eventually become the most produced American fighter of the Second World War's wartime production run.

Seversky, however, was by every contemporary account a far better pilot than he was a businessman. He was brilliant at demonstrating aircraft, charming at dinner, persuasive on a lecture stage, and genuinely innovative at the drafting table. He was not organized, not punctual with deliveries, and not interested in the administrative machinery that converts a promising aviation company into a financially viable one. The P-35 production contract for 77 aircraft was delivered chronically late, and the Army responded by canceling part of the order and buying Curtiss P-36 Hawks instead. When the company entered a new fighter competition in 1939 with the much-improved AP-4, which Seversky himself designed and which demonstrated impressive high-altitude performance, the contract again went to Curtiss and its XP-40. The competition on January 25, 1939 at Wright Field also included the Lockheed XP-38 and the Bell XP-39, and the XP-40 did not outperform either the AP-4 or the XP-38 at altitude. The Army chose it anyway, and the reason was purely practical: the XP-40 had been built on the tenth production P-36A airframe, meaning Curtiss already had the tooling and the production line running and could deliver aircraft to squadrons within months. The AP-4 and the XP-38 would both have required new facilities and new tooling. At $24,566 per aircraft, the P-40 was also dramatically cheaper than its competitors. With Europe already at war and the Army desperate for fighters it could actually have rather than the best fighter it could eventually build, production readiness beat performance across the board. The Army simultaneously ordered 13 YP-38s and 13 YP-39s for further evaluation, and also ordered 13 AP-4s as the YP-43 Lancer, a clear signal that it recognized the AP-4's high-altitude capabilities even while buying the P-40 in bulk. The Army did not reject the AP-4's performance. It rejected its delivery timeline. By April 1939, the Seversky Aircraft Corporation had accumulated losses of $550,000 and was falling further behind on its remaining production commitments with no credible plan to catch up.

The board of directors, led by financier Paul Moore, had reached a conclusion that General Henry "Hap" Arnold, Chief of the Army Air Corps, had already reached independently. Arnold had no legal authority over a private company's board of directors. What he had was something more powerful in practical terms: he controlled every production contract the Army Air Corps awarded, and Seversky Aircraft's entire revenue depended on those contracts. Arnold could not fire Seversky. He could, however, make clear to the board that the Army's future procurement decisions would reflect its confidence in the company's management and its ability to scale production. According to Air and Space Forces Magazine's published account of the period, Arnold wanted Seversky out of Seversky specifically because he did not believe the company could meet the production demands that the coming war would require. The board needed no direct order. They needed their contracts to keep coming, and they understood what Arnold's assessment meant for those contracts if nothing changed.

In May 1939, while Seversky was abroad on a European sales tour, the board removed him from the post of president. By October of that year, the company was formally reorganized as Republic Aviation Corporation. The name was a deliberate erasure. The board's reasoning, documented in contemporary accounts, was straightforward: the Seversky name had become associated with late deliveries, cost overruns, and eight years of unbroken financial losses, and they wanted a clean break from that reputation. "Republic" carried no such baggage, projected a patriotic American identity at a moment when government contracts were the company's entire lifeline, and removed any trace of the man the board had just ousted from the signage and the stationery. No source identifies whose specific suggestion the name was or what alternatives were considered. The board knew what they were walking away from. Where they were walking to mattered less than the distance from where they had been.

The man they chose to run the new Republic was not a random outside appointment. W. Wallace Kellett was a Princeton graduate who had driven ambulances for the American Field Service in France during the First World War before flying with the French air force. He had entered the aviation manufacturing business in 1919 and co-founded the Kellett Autogiro Corporation in 1929 with his brother Rodney Kellett, pioneering rotary-wing aircraft in the United States. His company's KD-1 autogyro, a two-seat open-cockpit machine powered by a 225-horsepower Jacobs radial engine with a three-bladed rotor that first flew in 1934, was purchased by the Army in 1935 and designated the YG-1, making it the first practical rotary-wing aircraft used by the United States Army. The same aircraft later inaugurated the first scheduled rotary-wing airmail service, flown by Eastern Airlines between Philadelphia and Camden, New Jersey, on July 6, 1939. Critically, he had been on the Seversky board since the company's founding in 1931, which meant he had spent eight years watching Seversky's management failures from the inside before being asked to fix them.

He wasted no time. Three weeks after taking the presidency he cut the payroll from 500 employees down to a working force of 185, then rebuilt it systematically around production discipline rather than design ambition. By the end of 1939 the company reported its first profit. When Seversky had been removed, unfilled orders totaled less than $100,000. Within a year Kellett had secured a $10 million Swedish export order and a $3.5 million Army contract, with a total backlog of $13 million. The reorganization Arnold had wanted, and the board had executed, was already working.

Seversky himself received $80,000 on his way out and a lawsuit's worth of legal frustration on the way back in. He fought to regain control of the company he had founded, but the battle dragged on without resolution until September 1942. The company he had built on Long Island went on to produce nearly 16,000 aircraft during the war without him at the helm.

The irony is that Seversky, freed from the administrative responsibilities that had overwhelmed him, turned out to be genuinely more valuable working elsewhere. He wrote. He lectured. He lobbied. His 1942 book "Victory Through Air Power," published by Simon and Schuster, argued with urgent clarity for the strategic bombing campaign and the necessity of American long-range airpower. It was a Book-of-the-Month Club selection that sold broadly enough that a Bugs Bunny cartoon parodied it the following year as "Victory Thru Hare Power," which is as reliable an indicator of cultural saturation as any box office figure. Walt Disney read the book and felt its message was important enough that he personally financed a feature-length animated documentary of the same title, released on July 17, 1943, in which Seversky himself appeared on screen alongside the animated sequences. Disney's personal investment was $788,000, and the film barely broke even at $799,000 in box office receipts, leaving Disney carrying the deficit until 1945. Winston Churchill and Franklin Roosevelt screened it together at the Quebec Conference in 1943. Historians disagree about how much the film directly influenced Allied strategy, but it is documented that both men saw it and that Seversky's arguments for long-range airpower were finding receptive audiences at the highest levels of Allied command during the same period. On the strength of that book and the hundreds of lectures that accompanied it, Seversky almost certainly did more for American airpower policy from outside his company than he ever could have managed from within it. The Thunderbolt that bore the design DNA of his P-35 went on to be built in numbers that Seversky, had he remained at the helm, almost certainly could not have achieved.

Kartveli stayed.

The design lineage from the P-35 to the Thunderbolt runs through a series of increasingly ambitious experiments that Kartveli conducted with the turbosupercharger. The principle was straightforward enough: an exhaust-driven turbine compresses outside air and forces it into the engine at high altitude, where the thin atmosphere alone cannot sustain full power output. The mechanical result was that a turbosupercharged engine could maintain sea-level performance, or something close to it, at altitudes where normally aspirated engines lost a third or more of their power. Kartveli had first installed a turbosupercharger in the AP-4, positioning it in the tail cone and running exhaust ductwork the length of the fuselage to drive it. The installation was complex, requiring extensive ducting that consumed fuselage volume other designs devoted to fuel and structure. The Army found the performance at altitude sufficiently impressive to order thirteen YP-43 Lancers, which entered limited production and saw some service in China, where operating above 25,000 feet gave a meaningful advantage over the altitude ceiling of many Japanese aircraft, though the P-43 was generally unpopular with pilots due to fuel-tank vulnerability and maintenance difficulties. The P-43 was not a combat-worthy aircraft in most respects, but it taught Kartveli what he needed to know about integrating a turbosupercharger into a production airframe.

The Army Air Corps issued its circular proposal for a new lightweight interceptor in the autumn of 1939. The specification called for a small, fast aircraft built around the Allison V-1710 inline engine, the same powerplant going into the P-38 Lightning and P-40 Warhawk. The Allison was not an arbitrary or uninformed choice. It was, in 1939, the only American-designed liquid-cooled V-12 aircraft engine in existence. The Army Air Corps had first purchased a V-1710 prototype in December 1932 and had been investing in its development ever since, with the specific intent of providing a narrow-frontal-profile engine that would produce less drag than the large-diameter radials that dominated American fighter design. A slender inline engine theoretically meant a smaller, faster airframe, and in the mid-1930s that logic had seemed sound.

The flaw in the logic, which the Battle of Britain was beginning to expose as the specification was being written, was in how the Army assumed high-altitude performance would be achieved. The Army's doctrine relied on turbo-superchargers bolted externally to the airframe to compensate for the Allison's limited mechanical supercharger development, the same approach Kartveli had been applying since the AP-4. This worked in the P-38, where two Allisons with turbo-superchargers gave the Lightning genuine high-altitude capability, but in a single-engine design the turbocharger's bulk, weight, and ducting consumed enough of the airframe's volume and weight budget that the inline engine's drag advantage largely disappeared. The British Rolls-Royce Merlin had taken the opposite approach, developing a two-stage, mechanically driven supercharger integrated into the engine itself rather than bolted onto the outside of the airframe. That solution, which would eventually power the P-51B Mustang to genuine high-altitude performance without the turbocharger's weight and plumbing penalties, was still in development in 1939 and not yet available to American designers. None of the American manufacturers specifying Allison-powered lightweight interceptors in 1939 knew how decisive the Merlin's internal supercharger solution would prove. They were working with what existed. Kartveli submitted a design he designated the AP-10, which the Army numbered the XP-47. Curtiss submitted the XP-46, a new design that incorporated features observed in European fighters while sharing the P-40's general layout, ordered in September 1939. Douglas submitted the XP-48, a small lightweight design powered by a Ranger inline engine, which the Army cancelled in February 1940 before a prototype was built after performance projections proved increasingly optimistic. North American Aviation, whose NA-73X prototype would eventually become the P-51 Mustang, was not a participant in this competition, though North American did purchase aerodynamic data on the P-40 and XP-46 from Curtiss for $56,000 during development of the NA-73X. All the companies that did compete were working the same basic concept: the smallest possible airframe wrapped tightly around an inline engine. This was, in hindsight, precisely the wrong direction to be heading.

The Battle of Britain was being fought while both designs were on the drawing boards. Combat reports from England were arriving in Army Air Corps offices throughout the summer and autumn of 1940, and what they described was an air war that had already rendered lightweight interceptor thinking obsolete. Fighters operating without armor plate behind the pilot were being killed by aircraft they never saw. Fighters without self-sealing fuel tanks were burning when cannon shells found them. The performance required to fight at 25,000 feet over Germany, where the strategic bombers that both the Army and the RAF were building would eventually need escort, demanded more than the Allison V-1710 could deliver even under ideal conditions. Adding armor and self-sealing tanks to a design that had been optimized for minimum weight meant the Allison could no longer get the result airborne fast enough to matter. By the spring of 1940, Kartveli and the Army both reached the same conclusion simultaneously: the XP-47 and the improved XP-47A were dead ends. The XP-46 investigation by Curtiss arrived at the same finding, and that program died without producing a prototype.

It was in the specific context of those cancelled designs, and the failure of the XP-44 concept that followed, that Kartveli made his appointment at Wright Field, went to Washington, and heard the Army's assessment of the XP-44's inadequacy in person. He boarded his train home that spring afternoon in 1940 and picked up his pencil. What he drew on the way back to Penn Station in Manhattan was not a refinement of the aircraft he had been working on. It was a different kind of statement about what a fighter actually needed to be. The Cradle of Aviation Museum's account of the development states that when Kartveli arrived at Penn Station, according to later retellings of the story, he had the basic outline of the fighter that would eventually break the back of the Luftwaffe in 1944. No contemporary documentation confirms the precise timing of the design's conception, but the train journey from Washington to New York falls within the period when Kartveli would have begun the fundamental rethinking the Army's rejection of the XP-44 demanded.

The design Kartveli submitted to the Army Air Corps on June 12, 1940, began with two non-negotiable premises. The first was that the next American fighter needed enough armor, self-sealing fuel capacity, and structural integrity to survive in the combat environment that Europe had already defined. The second was that the turbosupercharger was not a feature to be accommodated within an airframe, but the defining constraint around which everything else in the design had to be organized.

The turbosupercharger Kartveli specified was the General Electric B-series unit. The B-2 variant used with the R-2800 engine weighed 135 pounds for the turbine assembly alone, with the complete installation including intercoolers, ductwork, waste gate, and associated plumbing adding several hundred pounds more to the total. At maximum engine RPM the R-2800 moved more than 200 pounds of air per minute, which gives a sense of the airflow volumes the turbocharger and its associated ducting had to handle. The unit was large enough, and generated enough heat in operation, that it could not be placed anywhere near the fuel, the pilot, or the engine without creating problems that no amount of insulation could solve. It went in the aft lower fuselage, as far from those three concerns as the airframe allowed.

The turbocharger and its associated ductwork were the first components Kartveli placed in the design. Everything else was built outward from them, in the same design logic Seversky had first applied to the AP-4. Intake air entered through a scoop built into the underside of the nose cowling, traveled forward to an intercooler that removed the heat of compression, then continued rearward through the main duct to the turbocharger positioned in the lower aft fuselage, and finally traveled forward again to the injection carburetor on the engine. The total distance that air traveled from the nose intake to the carburetor and back was forty-three feet. In several sections the main ducts were large enough for a man to crawl through. The energizing exhaust gases from the engine traveled nineteen feet rearward to the turbine wheel, protected from the intense heat at the exhaust exit shroud by a carefully engineered boundary layer of cool air. The intercooler dumped its waste heat through two exit doors in the sides of the fuselage aft of the cockpit, which is why P-47s in photographs often show what look like gill vents in the fuselage sides that no other single-engine fighter of the period shares.

All of this ducting, installed in the lower fuselage aft of the cockpit, produced the characteristic deep belly that defines the Thunderbolt's profile. The fuselage needed to be deep enough to house the turbocharger and its waste gate assembly. It needed to be long enough to accommodate the exhaust run from the engine. The intercooler air exits required fuselage panels that no lightweight interceptor would ever have included. The resulting airframe was large in a way that had nothing to do with poor design and everything to do with the physics of what Kartveli was trying to accomplish. Alexander Kartveli reportedly described the aircraft in progress to his colleagues as "It will be a dinosaur, but a dinosaur with good proportions." He was not wrong on either count.

The Pratt and Whitney R-2800 Double Wasp radial engine that Kartveli specified to take advantage of all that forced air was the same engine family powering the Vought F4U Corsair and the Grumman F6F Hellcat simultaneously, a fact that says something significant about the R-2800's dominance of American high-performance aviation in the early 1940s. The variants were not identical. The Corsair's R-2800-8 and later -18W relied on the engine's own single-stage mechanical supercharger without an external turbocharger. The Hellcat used a similar installation. The P-47's R-2800-21 and -59 variants were specifically configured to work in conjunction with the GE turbocharger system, the engine's own internal supercharger feeding into the turbocharger rather than handling high-altitude compression on its own. Three of the most capable American fighters of the war shared the same basic powerplant and extracted fundamentally different performance profiles from it depending on how the surrounding airframe managed the induction system.

The R-2800 was an eighteen-cylinder, two-row air-cooled design with 2,800 cubic inches of displacement, capable of producing 2,000 horsepower at its initial military rating and significantly more in later variants with water injection. It required a broad cowling shaped around the engine face in what engineers called a horse-collar ellipse, which also admitted cooling air for left and right oil coolers and the intercooler system, producing an inlet opening of unusual width and depth that gave the Thunderbolt its distinctive blunt nose profile.

Eight .50-caliber Browning M2 machine guns in the wings were a requirement Kartveli considered non-negotiable from the first proposal. A conventional landing gear could not provide enough ground clearance for a propeller large enough to absorb the R-2800's full power output without placing the gear so far outboard that the wing guns and their ammunition would have no room. The original three-bladed propeller on the XP-47B measured 12 feet and 2 inches in diameter. The later four-bladed paddle propeller that resolved the climb rate problem measured 13 feet and nearly 2 inches in diameter. Providing ground clearance for a propeller that size, on an aircraft that also needed eight wing-mounted guns with their associated ammunition bays, presented a geometric problem with no obvious answer.

The Vought F4U Corsair, which used the same engine family and a similarly sized propeller, had solved the ground clearance problem by bending its wing into an inverted gull shape, shortening the gear legs while maintaining the propeller's clearance above the ground. That solution worked for the Corsair because the Navy's carrier requirements drove the entire airframe geometry from the beginning. Kartveli was working from different constraints entirely. The P-47's wing was a straight cantilever design, and moving the main gear attachment points outboard to clear the gun bays left the wheel wells with insufficient depth for a conventionally extended strut. Three alternative solutions each broke something else. Making the struts longer and accepting a wider track pushed the gun bays further outboard and created structural problems the wing could not absorb. Reducing the propeller diameter to shorten the struts compromised the high-altitude performance the entire design existed to deliver. Moving the guns out of the wings violated the eight-gun requirement Kartveli had considered non-negotiable from the beginning. The telescoping shrinkage strut was not the first instinct. It was what remained after every other option had been eliminated by the constraints the design had already imposed on itself.

Kartveli had been thinking about landing gear geometry since the P-35, whose retractable gear had been criticized for being only marginally better than a fixed arrangement at reducing drag. The XP-47B's solution drew on that accumulated thinking applied to a more extreme constraint than any previous Republic design had encountered. When the pilot raised the gear, hydraulic pressure triggered the piston within the main shock strut to compress into a bottomed-out position, with the displaced air redirected into an auxiliary chamber above it. The strut physically shortened by nine inches during the retraction cycle, collapsing inward so the gear could fold into a wheel bay sized for the shorter retracted length. When the gear came back down, the strut extended to its full length under hydraulic pressure before the wheels reached the ground.

A pilot or mechanic looking at the gear in its down position would see no obvious evidence of this mechanism. The nine inches of extra length was already fully deployed. The engineering was invisible until the gear started moving. Two designers working from the same engine family and the same ground clearance problem reached completely different structural solutions. One bent the wing. The other compressed the strut. The Corsair's answer was visible in its silhouette from a mile away. The Thunderbolt's answer was invisible until the gear started moving.

The resulting aircraft was, as Kartveli had promised, enormous by fighter standards. It was also, when the performance calculations were checked and rechecked, going to be very fast at very high altitude. The Army Air Corps had asked for a lightweight interceptor. What it received instead was the largest and heaviest fighter aircraft that had yet been proposed to any government anywhere. The Army, studying the performance projections and recognizing that the Battle of Britain had already made lightweight interceptor thinking a historical curiosity, ordered a prototype on September 6, 1940. One week later, before a single component of the prototype had been machined, the Army ordered 171 production P-47Bs and 602 P-47Cs in a single contract totaling $56,499,924. There was no production P-47A. The XP-47A designation had belonged to the stripped-down version of the original lightweight Allison-powered XP-47, ordered without military equipment to allow faster testing before the fully equipped version was ready. When Kartveli abandoned both and started over with the R-2800 design, both the XP-47 and XP-47A were cancelled and the new prototype took the next available letter, becoming the XP-47B.

The fact that both variants were ordered simultaneously before the prototype flew reflects a wartime procurement logic that had nothing to do with confusion and everything to do with urgency. What specifically distinguished the B from the C would only become clear through the prototype testing that was still ahead. What the Army cared about in September 1940 was getting production moving. The Army already knew the B model would need refinement before it reached its full potential. The improvements that would become the C were identified in advance. But incorporating those improvements required retooling, redesign work, and time that the war could not wait for. Rather than hold production until the better variant was ready, the Army ordered the B to get Republic's factory moving, the workforce trained, the supply chain established, and aircraft in pilots' hands as quickly as possible. Waiting for perfection while the enemy was already flying meant losing the war one uncontested mission at a time.

The prototype XP-47B emerged from the Republic factory at Farmingdale in the spring of 1941 and immediately attracted attention from everyone who came to see it. It sat significantly higher off the ground than any other fighter. The horse-collar cowl framed an engine that appeared disproportionately large even for an aircraft of the Thunderbolt's dimensions. The pilot's cockpit, perched high on the razorback fuselage, provided adequate view forward and to the sides. It provided almost none to the rear, a limitation that would shadow the aircraft through its first two years of combat. Inside, however, pilots who climbed up into the seat found it unexpectedly hospitable, "like a lounge chair" in the phrase that circulated among early test pilots. The cockpit was roomy, well organized, and comfortable in a way that smaller fighters were not. The aircraft's designers had clearly thought about what a pilot needed to spend several hours at high altitude, which was exactly the mission for which the Thunderbolt had been built.

On May 6, 1941, Republic test pilot Lowery Lawson Brabham taxied the XP-47B to the grass field at Farmingdale and opened the throttle. The aircraft lifted off after approximately 2,500 feet of ground roll, which Brabham noted as shorter than expected for a machine of that weight, though the run itself exposed the torque problem that would characterize the Thunderbolt throughout its production life. The R-2800's 2,000 horsepower turning that large propeller generated a pronounced left-yawing tendency on the takeoff roll that required constant right rudder to counter, and the long ground run before the rudder developed enough authority to control the swing meant that new pilots needed careful coaching before their first solo departure. Every P-47 variant would share this characteristic. When the paddle-blade propeller arrived on the D model and the diameter increased to over thirteen feet, the handbook added the requirement to keep the tail down longer than felt natural to prevent the propeller tips from striking the ground during rotation.

Almost immediately after liftoff, as Brabham climbed and the ambient pressure dropped around the exhaust ducting, smoke began filling the cockpit. The source was an oil drip in the exhaust system that the reduced ambient pressure was drawing into the cockpit space through the ducting seals. Brabham tried to open the canopy, which on the prototype was a hinged side door rather than the sliding panel that would appear on production aircraft. The door would not open in flight against the airstream. He tried the small vent window instead and discovered that the low-pressure zone around the vent drew more smoke in rather than less. He made a quick calculation, diverted to the paved runways at Mitchel Field a short distance away rather than landing back on Farmingdale's wet sod, and put the airplane on the ground. The landing was smooth. The brakes worked. The gear came down correctly. Army and Air Corps personnel who had heard a prototype of the new Republic fighter was operating in the area poured out of hangars to meet it as it rolled to a stop with its enormous propeller winding down. Nothing remotely like the XP-47B had ever landed at Mitchel Field. When Brabham climbed out he told the assembled crowd, "I think we've hit the jackpot."

The Army assigned the XP-47B back to Republic rather than taking it to Wright Field for government testing. The reason was practical: the prototype was the most thoroughly instrumented aircraft in the program and the fastest platform for evaluating engineering solutions to problems as they emerged in the early production aircraft. Every fix that Republic developed for the fabric control surfaces, the ignition system, the canopy, and the control loads needed to be tested on a flying aircraft before it could be committed to the production line. The prototype was that testbed. It did not sit in a hangar between May 1941 and August 1942. It flew continuously as a development tool, accumulating data that went directly into improving the aircraft that was already rolling off the Farmingdale line.

What the P-47 did not share with many of its contemporaries was a tendency to drop a wing on approach and landing. The flight manual described it as settling quickly but predictably when it lost flying speed. The wide-track landing gear, a direct consequence of the wing gun requirement pushing the gear outboard, gave the Thunderbolt substantial lateral stability on the ground. The same design decision that had created the engineering challenge of the telescoping shrinkage strut had accidentally produced an aircraft that was considerably more forgiving on landing than its size suggested.

The engineers were working through a longer list of problems. At high altitudes the ignition system arced, interrupting engine operation at the moment it was most needed. Control loads on the ailerons became severe enough at high speed to lock them in place. The hinged canopy door was clearly inadequate and needed to be replaced with a jettisonable sliding panel.

The most serious problem announced itself with a fatality. The first four production P-47Bs, serials 41-5896 through 41-5899, were delivered to the Army on March 18, 1942, and immediately entered testing programs run simultaneously by the Army at Wright Field, NACA at its Langley wind tunnel facility, and Republic at Farmingdale. Active duty Army pilots flew alongside Republic's test pilots given the factory's proximity and the urgency of getting the aircraft evaluated quickly. Army Lieutenant George Burrell was flying 41-5899 on March 26, 1942, just eight days after delivery, as part of that evaluation when the aircraft went out of control in a high-altitude dive and the tail assembly failed as the fabric control surfaces tore. Burrell bailed out but the aircraft had lost so much altitude in the dive that his parachute had insufficient time to fully deploy. The accident grounded all P-47Bs while engineers investigated the cause.

The investigation revealed a problem that took some explaining, because fabric control surfaces were standard practice on virtually every combat aircraft of the period and the reasons for using them were sound engineering rather than penny-pinching. The primary reason was flutter prevention through balance. A control surface must be statically balanced around its hinge line, with counterbalance weight ahead of the hinge to prevent destructive vibration at high speed. The lighter the control surface itself, the less counterbalance weight is required. Switching from fabric to metal skin adds weight behind the hinge, which requires substantially more lead counterbalance forward of it, compounding the weight penalty. Fabric provided more than a double weight saving when the full balance requirements were accounted for, and at the tail surfaces where any excess weight has the greatest effect on handling, that saving was meaningful. The P-51 Mustang flew its A, B, and C models with fully fabric-covered control surfaces at the same altitudes and comparable speeds as the P-47, and the P-51D retained fabric on its rudder and elevators through the entire war. The Corsair served throughout the war with fabric control surfaces. Speed and altitude alone did not explain what had killed Burrell.

What the investigation pointed to was the specific aerodynamic environment around the P-47's tail surfaces. The aircraft's deep belly, the complex turbocharger ducting running the length of the lower fuselage, and the exhaust system created airflow characteristics around the tail that were unlike those of conventional fighters. In the high-speed dives that the Thunderbolt's performance naturally invited, those conditions produced pressure differentials between the interior of the fabric-covered surfaces and the outside air that the fabric could not withstand. The fabric ballooned outward, distorting the aerodynamic shape of the control surface, and at sufficient speed it tore, taking control authority with it. It was not that fabric was generically inadequate at the P-47's speeds. It was that the P-47's specific airframe configuration created local conditions at the tail that made fabric unworkable in that particular installation, while other aircraft flying at similar speeds with conventional fuselage shapes did not encounter the same problem to the same degree. The answer was all-metal control surfaces, which neither ballooned nor tore regardless of what the surrounding airflow was doing.

The prototype was eventually lost on August 5, 1942, when civilian test pilot Fillmore "Fil" Gilmer, a 31-year-old former naval aviator and Huntington resident, took it aloft for a test flight. During the climb, Gilmer interrupted the landing gear retraction cycle, leaving the tailwheel sitting in the supercharger's exhaust stream. The heat ignited the tire, which then set the magnesium turbocharger hub alight. When the burning assembly retracted into the fuselage it severed the tail control rods. At between 11,000 and 12,000 feet, Gilmer felt the stick go light and useless. With no pitch control and the aircraft entering a near-vertical plunge, he had sufficient altitude to bail out, unlike Burrell, who had been in a dive and lost too much altitude before he could clear the aircraft. Gilmer survived. The XP-47B went into Long Island Sound off Eatons Neck. The loss was not publicized at this early stage of the war. The wreck, including the turbocharger that had caused the fire, was discovered by recreational diver Kirby Kurkomelis in 2012 in approximately fifty feet of water. Before that accident, the prototype had achieved 412 miles per hour at 25,800 feet in level flight and had climbed from sea level to 15,000 feet in five minutes. Those numbers confirmed what the design calculations had promised. At high altitude, with the turbosupercharger doing its job, the XP-47B had been very fast.

Republic worked through the list with systematic determination. All-metal control surfaces replaced the fabric ones. The ignition system was pressurized to prevent arcing at altitude. Balance panels reduced the rudder loads to manageable levels. The jettisonable sliding canopy replaced the hinged door. Each correction required testing and each test produced new data. The production P-47B that emerged from this process was significantly better than the prototype. The Army, which had already committed to 171 of them before the prototype flew, received them beginning in late 1941 and early 1942 with reasonable satisfaction.

The P-47Bs that came off the Farmingdale line went first to the 56th Fighter Group for training and operational evaluation. No P-47Bs went overseas for combat operations. When the 56th sailed for England on January 6, 1943, aboard the RMS Queen Elizabeth, it left its P-47Bs behind entirely. A new stateside group took over the worn B models. The P-47Cs that would equip the 56th in England had been shipped separately, crated and loaded onto cargo vessels, and reassembled at British facilities by the British Reassembly Division before the group arrived. The Queen Elizabeth carried pilots, ground crews, and equipment. It did not carry aircraft. When the 56th's personnel reached Kings Cliffe, brand new P-47Cs were already waiting on the ramp. Colonel Hubert "Hub" Zemke, who commanded the 56th, had expected to continue flying the aircraft his group had spent months learning. Instead he found an entirely fresh batch of the improved variant, assembled by British ground crews, at a base his pilots had never seen. The group spent several more weeks in additional training before VIII Fighter Command considered it ready for combat. The 171st and last P-47B was retained as a test platform, redesignated the XP-47E, and used to evaluate the R-2800-59 engine, a pressurized cockpit, and a new Hamilton Standard propeller. It was, in effect, the B model's final contribution to the aircraft it had helped develop.

The P-47B itself differed from the prototype in three critical ways: it used the production R-2800-21 engine rather than the experimental powerplant, it replaced the hinged prototype canopy door with a sliding panel that could be jettisoned in an emergency, and it substituted all-metal ailerons and elevators for the fabric-covered surfaces that Lieutenant Burrell's accident had proved were fatally inadequate at high speed. The P-47C refined those corrections further, adding an upgraded General Electric turbosupercharger regulator and a strengthened tail assembly. The C-1 production block added an eight-inch fuselage extension forward of the cockpit to correct a center of gravity problem identified in B model testing and to ease engine maintenance access. The C-2 block added a centerline hardpoint capable of carrying either a 500-pound bomb or a 200-gallon external fuel tank, giving the Thunderbolt its first meaningful range extension beyond its internal fuel supply and planting the seed of the fighter-bomber capability the aircraft would eventually make its own.

By early 1943, three Eighth Air Force fighter groups were equipped with Thunderbolts and working toward combat readiness. The 56th was at Kings Cliffe learning its new P-47Cs. Two groups already stationed in England had been given Thunderbolts as well: the 4th Fighter Group, built around the core of experienced American volunteers who had flown with the Royal Air Force Eagle Squadrons before the United States entered the war, and the 78th Fighter Group, which had trained on the Lockheed P-38 Lightning and was not enthusiastic about trading it for something else.

Neither the 4th nor the 78th received the Thunderbolt with enthusiasm. The 4th's pilots had flown the Spitfire, a superb aircraft with delicate responsive handling and excellent visibility in all directions, and the contrast with the massive, heavy Thunderbolt was not flattering. Colonel Don Blakeslee of the 4th, surveying his new mount, reportedly said of the Thunderbolt, "It oughta dive, it sure can't climb." The 78th wanted their P-38s back.

It was around this time that the aircraft acquired the nickname that would follow it through the rest of the war and into history. The 4th Fighter Group's pilots, deeply unhappy about surrendering their Spitfires for what James Goodson called "a seven-ton milk bottle," began calling it the Jug. The most likely origin of the name is the chamber pot, known in period slang as a thunder mug or jug, which the aircraft's rotund profile apparently suggested to pilots who were not yet inclined to be charitable about it. The milk jug explanation, referencing glass milk containers of the era that shared the fuselage's portly silhouette, became the more widely circulated story after the war, and the juggernaut explanation, suggesting Jug was short for the unstoppable force the aircraft represented, appears to have originated in British aviation publications in the 1960s rather than from the pilots themselves. The Thunderbolt would go on to accumulate more nicknames than virtually any other Allied aircraft of the war, among them Big Ugly, Bucket of Bolts, Cast-Iron Beast, Repulsive Scatterbolt, T-Bolt, and Thunder Mug. Most of them were coined by pilots who would later describe the aircraft as the best machine they ever flew. The 56th, which had spent months learning what the aircraft was actually capable of, had a different view from the beginning. They knew the Thunderbolt was not a Spitfire below 15,000 feet. They also knew that the Thunderbolt's combination of turbosupercharged high-altitude performance and diving speed gave it a significant advantage over most Luftwaffe aircraft above 25,000 feet, particularly in the diving attacks that Zemke's tactics were built around.

The first combat mission took place on March 10, 1943, when the 4th Fighter Group flew a sweep over France. No enemy aircraft were encountered. The radios in the P-47s, it emerged, were entirely inadequate for the electromagnetic environment over occupied Europe, creating interference that made coordinated flight nearly impossible. Every radio was replaced with British equipment before missions resumed on April 8. On April 15, 1943, Major Donald Blakeslee of the same 4th Fighter Group that had been skeptical about the aircraft went looking for a fight and found one, bouncing a Focke-Wulf Fw 190 and shooting it down. Whatever Blakeslee thought about the Thunderbolt's climb rate, it had just produced the aircraft's first aerial victory. He would later become one of the most decorated fighter pilots of the European war, though he would eventually leave the P-47 for the P-51.

The problems the Thunderbolt's pilots encountered in early combat were real and deserve to be stated plainly. The aircraft's range in its initial configuration, even with auxiliary drop tanks that arrived in small quantities through the summer of 1943, was limited enough that it could not accompany bombers across Germany. The Luftwaffe understood this perfectly. German fighter controllers tracked American bomber formations on radar, massed their interceptors inside Germany, and waited at comfortable altitude for the P-47s to reach the limits of their fuel and turn back for England. On August 17, 1943, the Eighth Air Force launched its raid against the Messerschmitt factories at Regensburg and the ball bearing plants at Schweinfurt. P-47 escorts were able to accompany the bombers only as far as the Belgian border, roughly an hour's flying time from either target. When the Thunderbolts turned back, the Luftwaffe was waiting. Sixty B-17s were shot down, most of them over the unescorted portions of the route. LeMay, who led the Regensburg force and watched the P-47 escort depart while still over Belgium, is said to have commented afterward that the only fighters he saw over the target had black crosses on their wings. The second Schweinfurt raid on October 14, known ever after as Black Thursday, cost another sixty bombers and forced the Eighth Air Force to suspend deep penetration missions into Germany until the range problem could be solved.

In a turning fight at medium or low altitude, a Thunderbolt pilot who engaged a Focke-Wulf Fw 190 or a Messerschmitt Bf 109 on equal terms was, as pilots who tried it discovered, at a disadvantage. The Luftwaffe's standard approach against the P-47 was to attack from directly below and behind, where the razorback fuselage cut off the American pilot's rearward vision almost completely. The razorback design deserves some explanation, because it was not simply an oversight that designers allowed despite knowing it created a blind spot. The high spine fuselage behind the cockpit served genuine engineering purposes. It contributed to lateral stability in the roll axis, acting as a structural keel that resisted unwanted rolling motion and made the aircraft fly more predictably in straight and level flight. It was significantly easier and cheaper to manufacture than the blown Plexiglas bubble canopy that would eventually replace it, and early in the war the tooling and materials required to produce large one-piece canopies at scale simply did not exist. Pre-war tactical thinking had also assumed that fighters would primarily attack bombers rather than engage in the close turning combat where rear quarter awareness became survival critical. When the actual air war revealed that German fighters were the primary threat and that they would attack from the blindest possible angle, the razorback's liability became undeniable. The designers had not been careless in 1940. The combat environment had changed around them. It was in many respects the most exploitable single weakness the aircraft had.

What the Thunderbolt could do was absorb extraordinary punishment and continue flying home. On June 26, 1943, Second Lieutenant Robert S. Johnson of the 56th Fighter Group had his P-47 surprised by an Fw 190. According to Johnson's postwar account, which is the primary source for the specific figures, twenty-one 20mm cannon shells struck the aircraft along with more than 200 7.92mm machine-gun rounds, though exact hit counts vary among sources since they derive from postwar recollection rather than forensic wartime documentation. The instrument panel was destroyed. Fire flickered in the cockpit. The canopy shattered. Johnson tried to bail out, managed to force the jammed canopy open six inches, and could not squeeze through with his parachute attached. His landing gear dropped to the extended position on its own. Unable to bail out and unable to retract the gear, he flew the crippled aircraft back across the Channel at maximum throttle, trailing smoke, his windscreen coated in oil, and landed at the first airfield he could find. Mechanics counted more than 200 individual punctures in the airframe. The aircraft was repaired and flew again. Johnson, who described himself later as "not hurt worse than playing football," was back in the cockpit a week later. He would finish the war with 27 confirmed aerial victories, all in P-47s, the second-highest total of any American pilot in the European Theater.

The German pilot who attacked Johnson that afternoon had done everything correctly and it had not been enough. The assessment that circulated among Luftwaffe fighter pilots, reflected in postwar interviews and memoirs, was consistent: shooting down a Thunderbolt required more ammunition, fired more accurately, at closer range, than shooting down anything else the Americans were flying. The aircraft simply refused to die from what should have been fatal damage.

The tactical story of the 56th Fighter Group through 1943 is substantially the story of Zemke determining what the aircraft was actually suited for and building tactics around those characteristics rather than fighting the airplane's inherent nature. He was aggressive, analytical, and temperamentally suited to a command that required constant tactical improvisation against a capable and experienced enemy. He understood early that asking a Thunderbolt to turn and fight at low and medium altitude against German fighters was the wrong use of the aircraft. The Thunderbolt was a high-altitude platform with superb diving speed, devastating firepower, and the structural integrity to execute and survive maneuvers that would destroy lighter aircraft. He built his "Zemke Fan" tactic around those qualities, spreading his group into a wide formation that ranged above and ahead of the bomber stream, hunting from altitude and diving through German formations at the speed the Thunderbolt did best rather than maneuvering on their terms. He also trained his pilots with such intensity that Eighth Air Force headquarters called to ask why the 56th was consuming dramatically more fuel than other groups. Zemke told them it was fuel well spent.

The results were not subtle. The 56th Fighter Group finished the war credited by the Air Force Historical Research Agency with 665.5 aerial victories, the highest total of any fighter group in the Eighth Air Force, and produced two of the European Theater's most celebrated aces from the same squadron and the same aircraft.

Francis "Gabby" Gabreski was born Franciszek Stanislaw Gabryszewski on January 28, 1919, in Oil City, Pennsylvania, the second child of Polish immigrants whose father had anglicized the family name and operated a market, working twelve-hour days. Gabreski grew up in a household where hard work was assumed and education was the mechanism by which the second generation would exceed the first. He attended the University of Notre Dame in 1938, came close to failing out his freshman year, and found his focus when he discovered flying. He took civilian lessons, enlisted in the Army Air Corps reserve in 1940, completed his wings, and arrived in England in October 1942 with the 56th Fighter Group.

What set Gabreski apart from most American pilots arriving in the theater was what happened next. Because of his Polish heritage and his fluency in the language, Gabreski requested assignment to a Polish fighter squadron flying with the Royal Air Force. His request was approved, and he was assigned to No. 315 Squadron at RAF Northolt, where he flew the Supermarine Spitfire Mk IX on patrol sweeps over the Channel. He flew twenty combat missions with the Poles, encountered the Luftwaffe, and absorbed from experienced pilots who had been fighting since 1939 a tactical education that no amount of training back in the United States could have provided. He understood how German pilots attacked, how they broke off, what they did when they were surprised and what they did when they weren't. He brought two Polish pilots he had flown with at Northolt, including ace Squadron Leader Boleslaw "Mike" Gladych, into the 56th Fighter Group when he rejoined it.

In the 56th's 61st Fighter Squadron, flying Thunderbolts against the Luftwaffe, Gabreski applied what the Poles had taught him. He scored his first aerial victory in August 1943 and his fifth, making him an ace, in November. By March 1944 he had 18 aerial victories and was among the three highest-scoring American aces in the theater. Robert Johnson, flying in the same 61st Fighter Squadron, had already broken Captain Eddie Rickenbacker's World War I record of 26 victories on May 8, 1944, becoming the first American pilot in the European Theater to do so. Neel Kearby had already equaled and passed Rickenbacker's record in the Pacific by that point, but Johnson was the first to do it in Europe. On June 27, 1944, Gabreski scored his 27th kill, equaling Johnson's total and tying Rickenbacker's record simultaneously. On July 5, 1944, leading the 56th on a mission over Germany, he destroyed a Messerschmitt Bf 109 to reach 28 confirmed aerial victories, surpassing Johnson and setting a new record as the highest total recorded by any American pilot fighting in the European Theater of Operations. That record was never equaled by any other American pilot in Europe.

Fifteen days later, scheduled for rotation home, Gabreski spotted Heinkel He 111s parked on the airfield at Niedermendig, Germany, and led his flight down on a strafing pass. He came in too low. His propeller struck the ground. He survived the crash landing but was taken prisoner and spent the remaining ten months of the war in German captivity. He returned to the United States, married his fiancee, and came back to military service in time for Korea, where he shot down 6.5 MiG-15s to become one of only seven American pilots to achieve ace status in two wars.

The parallel story of those months, visible only in retrospect, is the evolution of the aircraft itself. Two deficiencies dominated the combat feedback coming back to Republic's engineering teams through 1943: limited range and an inadequate climb rate. A third, the rearward visibility problem created by the razorback fuselage, was receiving constant complaints but no immediate engineering answer.

The range problem was addressed incrementally through larger external drop tanks, which arrived in useful quantities through the summer of 1943 and were progressively improved through the following year. By early 1944, when Eighth Air Force commander General James "Jimmy" Doolittle authorized American fighters to leave the bomber stream and actively hunt German fighters wherever they could be found, the P-47 had sufficient range to reach deep into Germany. In March 1944, Thunderbolts escorted bombers all the way to Berlin for the first time. The campaign that followed, in which American fighters swept German airfields and attacked Luftwaffe interceptors wherever they appeared, decimated the Luftwaffe's experienced pilot strength in the spring of 1944 and measurably reduced Germany's ability to defend its airspace by the time Allied troops landed in Normandy in June. The range problem was managed rather than solved in the European Theater. It would not be fully solved until the P-47N, designed entirely for Pacific operations, introduced a fundamentally different wing.

The climb rate problem had a more specific engineering answer. The original production P-47B and C models flew with a three-bladed Curtiss Electric propeller spanning 12 feet and 2 inches in diameter. The R-2800 engine produced significantly more torque than that propeller could efficiently convert into thrust, particularly at lower altitudes and during climb. Beginning with the P-47D-22 production block at the Farmingdale facility, aircraft were delivered with a new Hamilton Standard Hydromatic propeller featuring four broad-chord blades spanning 13 feet and nearly 2 inches. The Evansville plant used Curtiss Electric paddle-blade props of the same basic diameter on its production runs. The wider blade chord grabbed substantially more air per revolution and transferred a greater fraction of the engine's power into forward motion during climb. Pilots who flew both propeller types reported a climb rate improvement of approximately 400 feet per minute, a meaningful gain in combat where altitude was simultaneously a tactical weapon and a survival mechanism. Earlier production aircraft in the field received the new propellers as retrofits. Pilots who had been complaining about the Thunderbolt's initial climb performance found themselves with a substantially different aircraft under them.

The visibility problem required something more fundamental. The razorback fuselage, which extended upward behind the cockpit and blocked nearly all rearward vision, was a structural element of the aircraft, not a component that could be exchanged at a depot. Modifying it required cutting down the entire aft fuselage structure and finding a canopy design that could provide 360-degree visibility above a dramatically lowered cockpit surround.

In 1943, a Republic engineer obtained a pre-production blown Plexiglas bubble canopy from the Hawker Typhoon program. The Typhoon had introduced this design to British service aircraft earlier that year, replacing the older car-door cockpit entry that had characterized early Typhoon production. Republic modified a P-47D-5 with the cut-down aft fuselage and the Typhoon-derived bubble canopy and designated it the XP-47K. The modification worked. A second P-47D-5 was simultaneously modified to evaluate an increase in internal fuel capacity, and designated the XP-47L. Both improvements were consolidated into production beginning with the P-47D-25 at the Farmingdale plant. Pilots called the result the bubbletop, and the nickname stuck. The cut-down fuselage eliminated the rearward blind spot that Luftwaffe pilots had been exploiting since the aircraft entered combat. It also reduced drag slightly and, combined with the additional internal fuel of the XP-47L configuration, produced what became the definitive Thunderbolt variant. The aircraft's silhouette changed so completely with the razorback spine gone that bubbletop and razorback P-47s photographed side by side look like they might be different designs. They are the same aircraft, separated by an engineering solution that arrived approximately two years later than the pilots who needed it would have preferred.

As the P-51 Mustang arrived in the European Theater in growing numbers through late 1943 and 1944 and took over the long-range escort role it was uniquely suited for, the Thunderbolt shifted toward a mission it was also uniquely suited for, though nobody had initially designed it with that mission in mind. The P-47 became a fighter-bomber, and at that role it proved exceptional.

By May 1944, thirteen of the Ninth Air Force's fighter groups were equipped with P-47Ds fitted with underwing hardpoints capable of carrying 500-pound bombs, later 1,000-pound bombs, and eventually the five-inch High Velocity Aircraft Rocket, known to pilots as the "Holy Moses." After the Normandy landings on June 6, 1944, Thunderbolt groups followed the Allied breakout from the beachhead, operating from hastily constructed airstrips in France and conducting what amounted to a systematic destruction of the German Army's ability to move by road and rail in daylight. A three-second burst from a Thunderbolt's eight .50-caliber guns sent approximately 340 rounds downrange, a figure that accounted for trucks, light armor, and locomotive boilers with equal efficiency. For heavier targets, the rockets and bombs were available. For Panther and Tiger tanks with their thick deck armor, concentrated strafing passes against the engine compartments and road wheels, combined with rocket attacks, could stop vehicles that direct .50-caliber fire could not penetrate.

During Operation Cobra in late July 1944, the American breakout from Normandy's hedgerow country, P-47 groups caught a German armored column trapped near Roncey and attacked it continuously from 3:10 in the afternoon until 9:40 that evening. Ground investigation afterward found 66 tanks, 204 vehicles, and 11 artillery pieces destroyed in that single day's work. It was not an isolated incident. It was representative of what the Thunderbolt was doing across the length of the German retreat through France and Belgium.

From the Normandy landings on June 6, 1944, through Germany's surrender on May 8, 1945, Thunderbolt pilots flying from England and the Continent destroyed or damaged 6,000 enemy tanks and armored vehicles, 68,000 trucks, 9,000 locomotives, 86,000 railroad cars, and 60,000 horse-drawn vehicles, according to the Smithsonian Institution's compiled statistics. The aircraft flew more than 545,000 combat sorties during the war and dropped more than 132,000 tons of bombs. Those numbers represent one of the most significant contributions to the destruction of Germany's ground transportation infrastructure of any single aircraft type in Allied service, though direct comparison across all Allied aircraft types is difficult given the complexity of the campaign statistics. The Eighth Air Force's strategic bombers eventually destroyed the Ruhr industrial complex. The Thunderbolt, operating at tree-top level against things that moved, helped ensure those industries had no way to get their products to the front.

The question of why the 56th Fighter Group was the only Eighth Air Force group that kept its P-47s when every other group transitioned to P-51s has a more complicated answer than group pride. Zemke himself, by most accounts, recognized the P-51's merits and was not ideologically committed to the Thunderbolt. The decision was made in January 1944 while Zemke was temporarily away from the group, and by the time he returned it had become institutional fact. What is certain is that the 56th's pilots had spent more time mastering the Thunderbolt's specific strengths than any other group in the theater, and those pilots had produced results that made argument difficult. Zemke transferred command of the 56th in August 1944, volunteering to take over the 479th Fighter Group, which was converting from P-38 Lightnings to P-51 Mustangs. He thought the 479th needed experienced leadership and he wanted one more hard challenge. In the 479th he scored 2.5 more aerial victories, participated in one of the first Allied shoot-downs of an Me 262, and was ordered to headquarters as chief of staff with more than 450 combat hours to his credit. He decided to fly one more mission before taking the desk job. On October 30, 1944, he ran into severe unforecast turbulence over Germany. His P-51 lost a wing. Zemke bailed out, was captured, and ended the war as the senior Allied officer at Stalag Luft I at Barth on the Baltic Sea, responsible for nearly 9,000 prisoners. He negotiated the camp's transfer to the prisoners before Soviet forces arrived, then worked to repatriate American POWs who the Soviets were inclined to move east rather than return west. He came home with 17.75 confirmed aerial victories in 154 combat missions and the kind of record that made postwar aviation writers run out of superlatives. Robert Johnson, who had flown under Zemke's command and who knew the P-51 as well as any pilot in the theater, said it plainly in a postwar interview: Hub Zemke put one in a dive and when he pulled out he ripped the wings off, and that was how he became a prisoner of war. The Thunderbolt, Johnson implied, would not have done that.

The 56th Fighter Group, the only Eighth Air Force group that kept its P-47s when every other group transitioned to P-51s, received one last development of the aircraft in early 1945. The P-47M had been developed in response to the V-1 flying bomb threat that began in June 1944, when the first pulse-jet-powered German cruise missiles started crossing the Channel at approximately 400 miles per hour and exploding over London. The Army needed fighters fast enough to catch them, and the standard P-47D at its normal operating altitude was not always up to the task. Republic fitted three P-47D airframes with the Pratt and Whitney R-2800-57 engine with a General Electric CH-5 turbosupercharger, capable of pushing output beyond 2,800 horsepower at altitude, and added underwing dive recovery flaps to allow controlled deceleration after high-speed pursuit without the aircraft entering an unrecoverable dive. The resulting XP-47M achieved 473 miles per hour in level flight at altitude, making it the fastest variant of the Thunderbolt ever built and among the fastest Allied piston-engine fighters of the war at high altitude. It was not the fastest Allied fighter overall at all altitudes. The Hawker Tempest and de Havilland Hornet both exceeded it at lower altitudes. What the M model offered was exceptional high-altitude speed combined with the Thunderbolt's established structural toughness, a combination that suited the Me 262 problem better than most alternatives. By the time 130 production P-47Ms reached the 56th Fighter Group at Boxted in early 1945, however, the V-1 campaign against England had effectively ended. The M model never flew its intended intercept mission. What it did instead was hunt the Messerschmitt Me 262, Germany's operational jet fighter, exploiting the M model's superior service ceiling and shallow dive speed to close with aircraft that could outrun it in level flight. The P-47M went on to distinguish itself in exactly the kind of mission its speed was suited for. On March 14, 1945, two Arado 234 jet bombers were shot down by 62nd Squadron P-47Ms near Remagen. On March 25, Major George Bostwick of the 63rd Squadron and his wingman dispatched two Me 262s on approach to Parchim airfield. On April 5, Captain Fahringer of the 63rd pursued a lone Me 262 that had just shot down a B-17, caught it in a turning engagement, and destroyed it. The 56th Fighter Group was credited with seven German jet aircraft destroyed in total before the war in Europe ended, flying an aircraft that had not existed when the jets it was hunting first entered service. The piston engine's last chapter was being written by the same group that had flown the Thunderbolt's first combat mission two years earlier.

While the 56th Fighter Group was building its European reputation through 1943 and 1944, the Thunderbolt was simultaneously fighting a completely different war on the other side of the world. The Pacific Theater asked something fundamentally different of the aircraft, pitted it against a different enemy with different tactics, and required a different kind of officer to make it work. To tell that story properly requires stepping back to September 1942 and an airfield in New York.

The 348th Fighter Group had been activated at Mitchel Army Airfield in New York on September 30, 1942, one of the first USAAF groups to be equipped with the Thunderbolt from the outset. The group trained through the winter of 1942 and 1943 at bases in Connecticut, Massachusetts, and Rhode Island on P-47s. Which specific variant they flew during that stateside period is not documented in surviving records. Every man in the group believed they were going to England when they boarded the Army transport Henry Gibbons at Weehawken, New Jersey on May 15, 1943. The ship went through the Panama Canal instead. The group reached Brisbane, Australia on June 14, 1943, and discovered their aircraft had not yet arrived. The P-47D-2-RE Thunderbolts were being shipped separately. On July 21, 1943, the escort carrier USS Barnes arrived at Bretts Wharf in Brisbane carrying a deck load of partially disassembled Thunderbolts, their propellers, ailerons, elevators, and rudders removed for transport. The aircraft were offloaded at Bretts Wharf and towed to Eagle Farm airfield, where RAAF ground personnel assembled them under the supervision of technical representatives from Pratt and Whitney, Curtiss Electric Propeller, and Republic Aviation, reinstalling each aircraft's major components before it could fly. The group spent the following weeks running in engines on local training flights before beginning combat operations over New Guinea.

The aircraft the 348th flew in the Pacific were early P-47D razorbacks. The question of how they differed from the P-47C that equipped European groups has a direct answer: in the early production blocks, barely at all. The first P-47Ds off the Farmingdale line differed from the P-47C-5 that preceded them mainly in minor turbosupercharger exhaust system refinements, additional cowl flaps for better engine cooling, and more extensive pilot armor protection. The Evansville plant's first 110 D models were, according to Republic's own production records, completely identical to P-47C-2s. The D designation tracked incremental improvements being applied block by block across both plants rather than marking a dramatic design departure. A P-47C and an early P-47D parked side by side were, in the words of one contemporary assessment, distinguishable only by their serial numbers. The significant visual and performance changes most people associate with the D model, the bubble canopy, the paddle-blade propeller, the additional fuel capacity and underwing hardpoints, came through successive production blocks across 1943 and 1944 rather than arriving all at once with the D designation.

Colonel Neel Kearby commanded the 348th and had arrived in Australia with that group in June 1943. He was thirty-two years old, born in Wichita Falls, Texas, and considered practically ancient by USAAF standards where most combat pilots were in their early twenties. He had used the stateside training period to develop P-47 doctrine around the aircraft's specific strengths, diving speed, high-altitude engine performance, and structural toughness, rather than trying to make it fight like the lightweight nimble aircraft it was not. He arrived already knowing the aircraft intimately and already convinced it could be made to work in any theater, including one it had never been designed for.

What he found in General George Kenney's Southwest Pacific command was a theater built around the P-38 Lightning that had little enthusiasm for the Thunderbolt. Kenney reportedly told Kearby directly to prove the P-47 was worth operating in his theater or take his group back home. Kearby's response was to challenge incoming P-38 commander Lieutenant Colonel George Prentice to a mock combat engagement, which Kearby won convincingly by exploiting the P-47's altitude advantage and diving speed against the P-38's low-altitude maneuverability. The demonstration bought the 348th its chance. Pacific Theater commanders remained skeptical about the aircraft's suitability against Japanese fighters, and their skepticism was not without foundation. The reputation of the Japanese Ki-43 Hayabusa and the Mitsubishi Zero for low-altitude maneuverability was well founded on years of combat experience in China and Southeast Asia. Any pilot who tried to turn with them at medium or low altitude was going to lose, regardless of what he was flying, and nothing about the P-47's size and weight suggested it would be an exception. The Thunderbolt's redeeming qualities in the Pacific were the same ones that had defined it in Europe: performance at altitude, diving speed, and the capacity to absorb damage that would destroy lighter aircraft.

The tactical approach Kearby developed was not invented from nothing. Energy fighting, gaining altitude, diving through an enemy formation at high speed, and using the accumulated energy to zoom climb back to altitude before the enemy could react, had been part of RAF fighter doctrine since at least 1940 and was well established in European combat by the time Kearby arrived in the Pacific. What Kearby did was apply it with systematic rigor to a specific aircraft against a specific enemy in conditions that made it uniquely effective. Japanese fighters were optimized for low and medium altitude combat. Their pilots were trained in the aggressive turning engagements at those altitudes where the Zero and Ki-43 excelled. The P-47's turbosupercharged engine gave it a decisive performance advantage above 25,000 feet that Japanese aircraft simply could not match, and a diving speed that nothing in the Japanese inventory could catch. Kearby's tactic exploited both simultaneously. He was not a creative genius in a vacuum. He was a rigorous tactician who understood his aircraft, understood his enemy, and had the discipline to train his pilots to fight the aircraft's battle rather than the enemy's.

On October 11, 1943, Kearby led a four-aircraft mission to reconnoiter Japanese airfields near Wewak, New Guinea, flying his P-47D razorback "Fiery Ginger," one of four successive aircraft he named after his red-haired wife Virginia, working through Fiery Ginger, Fiery Ginger II, Fiery Ginger III, and Fiery Ginger IV across his combat career. Wewak sat at the extreme range of the P-47D from the 348th's base at Port Moresby, meaning the entire flight arrived over the target already deep into their fuel reserves with barely enough to complete the reconnaissance and fly home. After completing the reconnaissance objective, the flight spotted a large formation of Japanese aircraft. All four pilots knew their fuel state. Engaging was already a gamble. Kearby ordered the attack anyway.

In approximately fifteen minutes of combat, Kearby shot down three enemy aircraft on his first diving pass. When he saw that the fourth member of his flight, Lieutenant Raymond Gallagher, was being chased by two Japanese fighters, Kearby reversed and attacked both of them, destroying both. The reversal consumed fuel that none of the four pilots had to spare. That Kearby made it at all, let alone destroyed both pursuers, was the act that earned the Medal of Honor. The total for the mission was six Japanese aircraft destroyed by Kearby personally, four Nakajima Ki-43 Hayabusa fighters and two Kawasaki Ki-61 Hiens, with his wingmen Captains Dunham and Moore accounting for three more Ki-61s between them for a mission total of nine Japanese aircraft destroyed without the loss of a single American aircraft.

Kearby entered the mission with three confirmed victories, all scored in the Pacific in the P-47D: a Mitsubishi G4M Betty bomber and a Nakajima Ki-43 Oscar escort on September 4, and a Mitsubishi Ki-46 Dinah reconnaissance aircraft ten days later on September 14. His six kills on October 11 brought his total to nine and made him the first P-47 ace of the Pacific Theater of Operations. They also set a United States Army Air Forces record for the most aerial victories in a single mission at that point in the war. The USAAF record was later broken in January 1945 when P-51 pilot William Shomo shot down seven Japanese aircraft in six minutes over the Philippines. The overall American record of nine in a single mission was set in October 1944 by Navy F6F Hellcat pilot David McCampbell. Kearby's gun camera ran out of film during the fight, and two additional kills that his wingmen witnessed could not be officially confirmed. Had they been, his single-mission total would have stood at eight. A postwar examination of Japanese records suggested that only two of the nine victories credited to the full mission could be confirmed against Japanese loss records, though three additional aircraft were recorded as damaged. The disparity between American claims and Japanese records was a consistent feature of Pacific air combat accounting on both sides, where the chaos of a turning fight made precise confirmation difficult for any nation's record-keeping system. Kearby received the Medal of Honor from General Douglas MacArthur, who presented it personally in his office in Brisbane on January 23, 1944. He was the first United States Army Air Forces fighter pilot in the war to receive it. Every previous Army Medal of Honor awarded to an airman in the war had gone to bomber pilots or crewmen. Kearby was the first to receive it for fighter combat. By the time he was shot down and killed over Wewak on March 5, 1944, flying Fiery Ginger IV, Kearby had 22 confirmed aerial victories, more than any other P-47 pilot in the Pacific Theater. The wreckage of Fiery Ginger IV lay in the New Guinea jungle for decades after the war. The tail fin was eventually recovered and is now on display at the National Museum of the United States Air Force alongside a full-sized replica of the aircraft.

The aircraft Kearby flew over New Guinea bore only a general resemblance to the aircraft Brabham had nursed through the smoke on its first flight three years earlier. Republic produced the Thunderbolt across two factory lines, at Farmingdale and at Evansville, Indiana, which was added to meet demand that a single facility could not satisfy. Curtiss-Wright operated a third production line that built 354 P-47Gs before returning to its own programs. The P-47D model alone accounts for 12,602 of the total production run, making it the most produced sub-variant of any American fighter in the war. Through all variants, from the prototype through the final P-47N-25 that rolled off the Evansville line in October 1945, a total of 15,683 Thunderbolts were built according to the Smithsonian Institution's accounting, which is the figure used by Britannica and most major reference sources. Some sources use 15,686, and others cite 15,636, discrepancies that reflect different methods of counting prototypes, the P-47G Curtiss-built variants, and accounting differences between the two production plants. The precise figure is disputed but the order of magnitude is not. That total made the P-47 the most produced American fighter of the wartime production period. The P-51 Mustang's production run continued through 1951 for export and post-war orders and edged slightly ahead of the P-47 in total units across its full production life, but within the years the war was actually being fought the Thunderbolt was built in greater numbers than any other American fighter.

The P-47N that closed the production run was the Pacific Theater answer to a range problem that had constrained the aircraft throughout its European career. The N model introduced a completely redesigned wing with a straighter leading edge and substantially more internal fuel than any previous variant, extending the practical range to more than 2,000 miles. The R-2800-77 engine drove the performance envelope higher while the redesigned wing improved high-altitude handling. P-47Ns arrived on Saipan in the spring of 1945 and began escorting Boeing B-29 Superfortresses on raids against the Japanese home islands, the long-range Pacific escort mission the aircraft had been specifically built to perform. A total of 1,816 were produced before the war's end terminated the larger orders that would have built thousands more.

The last Thunderbolt to leave American military service in the active inventory did so in 1949, when the Air Force phased it out as jet fighters filled the front-line inventory. The aircraft had been redesignated the F-47 in 1948, its pursuit designation replaced by the fighter designation that all American tactical aircraft received when the Air Force became an independent service. Air National Guard units flew the F-47 until 1953, when the aircraft was finally withdrawn from American military service entirely. The Corsair remained in active Marine Corps squadrons until 1955 and in the Naval Reserve until 1957, outlasting the Thunderbolt in American service by several years, a reflection of the Corsair's later production run and its proven utility in the Korean War ground attack role that the F-47 was never called upon to fill. When the Korean War began in June 1950, Lieutenant General George Stratemeyer, the theater air commander, requested that F-47s be sent to Korea. The request was denied. The supply chain for spare parts had been wound down too completely to support a combat deployment, and the Air Force had no appetite for rebuilding it when jets were available. The Thunderbolt, which had spent six years absorbing everything the Luftwaffe and the Imperial Japanese Army and Navy could throw at it, never flew in anger again under American colors.

Under other colors it flew for years. The Latin American air forces that had received P-47s through wartime lend-lease and postwar military assistance programs kept them in service through the 1950s and into the 1960s. Peru, Venezuela, Colombia, Ecuador, Chile, Bolivia, Nicaragua, Honduras, and El Salvador all operated the aircraft. Iran, Turkey, Yugoslavia, Italy, Nationalist China, and the Republic of China on Taiwan received examples as well. The Peruvian Air Force, the last military operator anywhere, retired its final Thunderbolts in 1966, a full twenty-five years after the prototype's smoke-filled first flight at Mitchel Field.

Alexander Kartveli did not stop working when the Thunderbolt production line went quiet. He had already begun designing the F-84 Thunderjet before the war ended, starting the work in 1944 and applying the same organizational logic he had used on every previous aircraft: identify the primary constraint, place it first in the design, and build the airframe around it. The F-84 entered serial production in 1946, the first straight-wing jet fighter to serve the newly independent United States Air Force, and flew 86,000 combat missions in Korea. Republic's engineers and the pilots who flew its products had a saying by then that the company built airplanes for armies rather than for aces, meaning aircraft designed to keep flying after taking damage that would ground anything lighter. The F-84 in Korea validated that philosophy in jet form the same way the P-47 had validated it in piston form over France.

Kartveli followed the Thunderjet with the swept-wing F-84F Thunderstreak and then with the F-105 Thunderchief, which began as an internal Republic project in 1951 and grew into a Mach 2 low-altitude nuclear strike aircraft capable of carrying a tactical nuclear weapon to targets inside the Soviet Union. The F-105 was the largest single-seat, single-engine combat aircraft ever built by any nation, a lineage that connects directly back to the XP-47B prototype that had dwarfed every other fighter at Mitchel Field in 1941. The Thunderchief flew more combat missions over Vietnam than any other aircraft in the American inventory during the Rolling Thunder campaign of 1965 and 1966, absorbing antiaircraft fire and surface-to-air missiles on strike missions into the most heavily defended airspace in the history of aerial warfare to that point. It brought pilots home from damage that should have killed the airplane. The spirit was recognizable to anyone who had seen the mechanics at the 56th Fighter Group counting punctures in Robert Johnson's P-47 in the summer of 1943.

Kartveli died in Huntington, New York, on July 20, 1974, at the age of 77. He had lived long enough to see the Fairchild Republic A-10 Thunderbolt II take its first flight in May 1972, a straight-winged, twin-engine ground attack aircraft built around the largest aircraft gun ever installed in a production aircraft, the 30mm General Electric GAU-8 Avenger rotary cannon. The A-10 was a Fairchild Republic design that postdated Kartveli's active involvement in the company, but it carried his design philosophy forward in its name and in its organizing principle. The A-10 carries the Thunderbolt name in its designation, connecting it to its ancestor as explicitly as anyone could. The philosophy that produced both aircraft is the same: identify the thing that matters most in the mission, build the airframe to carry and protect it, and trust that the physics will produce an aircraft unglamorous enough to work. Kartveli had called the P-47 a dinosaur with good proportions in 1940. The A-10, which its pilots affectionately call the Warthog, is the same kind of animal: large, unglamorous, built for a specific purpose, and extraordinarily effective at that purpose for a span of decades that more elegant designs have rarely matched.

In 1940, the thing that mattered most was a turbosupercharger and forty-three feet of ductwork it required to function. Kartveli built his airplane around that constraint and promised the Army a dinosaur with good proportions. Fifteen thousand six hundred and eighty-three of them were built. By the end of the war, the pilots who flew them over France and Germany and New Guinea had destroyed a greater quantity of the enemy's ground transportation and armor than the statistics of almost any other Allied aircraft type could match. They had done it at a loss rate of 0.7 percent per mission, the lowest of any American fighter in the theater. The dinosaur turned out to have very good proportions indeed. Of the 15,683 built, approximately 58 complete airframes survive worldwide as of 2025, according to the most comprehensive accounting available. Sources differ on how many remain airworthy at any given time, with estimates ranging from fewer than a dozen to somewhat more depending on which aircraft are currently flying and which are undergoing restoration. What is certain is that when one does appear at an airshow, it draws a crowd. An aircraft that was once built at the rate of several per day now stops people in their tracks on a ramp.