The F4U was one of the most succesful fighters of WWII. It achieved a claimed 11.3 to 1 'kill ratio' in the Pacific. The Corsair was fast, but had some handling problems, and early versions tended to 'bounce' upon landing because of their stiff undercarriage. Therefore it was at first restricted to land-based marine units, and equipped all USMC fighter units in the Pacific in the second half of 1943. After some modifications, it was used also for shipboard operations, but continued to require careful handling. The vast majority of F4U sorties (54470 out of 64051) were flown from land bases. The F4U had an inverted gull wing, a streamlined fuselage of circular cross-section and a big propeller. The cockpit was set well back, and the restricted view was a serious problem. 12571 were built, of which 4017 by Goodyear and 735 by Brewster as the FG and F3A, respectively.
Engines: 1 * 1490kW P&W R-2800-8
Wing Span: 12.50 m
Length: 10.16 m
Height: 4.90 m
Wing Area: 29.17 m2
Empty Weight: 4074 kg
Max.Weight: 6350 kg
Max. Speed: 671 km/h
Ceiling: 11245 m
Max. Range: 1640 km
Armament: 6*mg 12.7 mm
Engines: 1 * 2350 hp P&W R-2800-18W
Wing Span: 12.49 m
Length: 10.13 m
Height: 4.50 m
Wing Area: 29.17 m2
Empty Weight: 4238 kg
Max.Weight: 6592 kg
Max. Speed: 684 km/h
Armament: 6*mg 12.7 mm 2*b454 kg 1*b726 kg
An unusual element of the RFP was that the Navy vowed to consider designs with liquid-cooled engines, in contradiction with a policy settled in 1927 that required air-cooled engines for shipboard aircraft. From the viewpoint of naval aviators, liquid cooling systems had serious disadvantages: They were heavier, more vulnerable, and more difficult to maintain. But in the late 1930s, there was a growing conviction in international aviation circles, that radial engines presented a too high drag penalty. Liquid-cooled engines with their smaller frontal area could be installed in a more streamlined fuselage. Hence the option to accept a fighter built for such an engine, in practice the Allison V-1710.
This engine was indeed chosen by Bell for their entry in the competition: The Bell Model 5 Airabonita, virtually a P-39 Airacobra with tailwheel landing gear, a slightly larger wing, and a stronger structure. As in the P-39, the engine was placed amidships, over the wing. The pilot sat in front of the engine, with a long extension shaft passing between his legs to drive the propeller up front. A 23mm Madsen cannon (or a .50 gun) and two .30 guns were installed in the nose, the cannon firing through the hollow propeller hub.
There was more choice in radial engines: The older Pratt & Whitney R-1830, and the new the Wright R-2600 and Pratt & Whitney R-2800. These air-cooled radial engines had a larger frontal area than the V-1710, and thus generated more drag. For the R-2600 and R-2800 this was compensated for by their power: While the V-1710 was hoped to deliver about 1150hp, the R-2800 was expected to generate 2000hp and more, and the R-2600 1500hp. Radial engines were chosen by Brewster, Grumman, Vought and Curtiss. Grumman proposed a development of the F4F Wildcat, that would be powered by the R-2600 engine. Brewster, manufacturer of the F2A Buffalo that had been the US Navy's first monoplane fighter, offered designs with the R-2600 or R-2800. Curtiss proposed developments of the P-36 Mohawk, powered by either the R-2600 or the older R-1830 engine.
In April 1938, Vought proposed its two designs to the US Navy. One, called V-166A by Vought and "Vought A" by the USN, was powered by the R-1830. The other, the V-166B or "Vought B", was designed around the new Pratt & Whitney R-2800 Double Wasp radial engine. This was an 18-cylinder, two-row air-cooled radial. This engine would later also be installed in the competing Grumman F6F Hellcat and in the Republic P-47 Thunderbolt for the USAAF, but the new Vought fighter was the first to use this engine. The R-2800 later acquired a reputation as a powerful and very reliable engine. But it was also very bulky, and aircraft powered by it tended to be big.
In May 1938, the Bureau of Aeronautics evaluated the proposals. The "Vought B" was deemed to be the best one, with a merit figure of 86.4 on a scale from 0 to 100. Hence on 11 June, a contract was given for development of the Vought V-166B, the fighter that would become famous as the F4U Corsair.
The evaluation committee also recommended that the "Brewster A" proposal, rated third best, should be developed because of its alternative R-2600 engine. Because of the managament difficulties of Brewster, this never happened. Grumman received a contract to develop to F4F-3 version of the Wildcat, and won the simultaneous competition for a twin-engined fighter with F5F Skyrocket. Their R-2600 engined fighter was rejected, but in June 1941 the Navy would nevertheless order two prototypes of the F6F Hellcat --- which switched to the R-2800 during development. The Navy was also sufficiently intrigued by the Bell proposal to order a prototype, named the XFL-1. But the Bell fighter, ranked sixth of the competitors, was obviously not destined to enter production, and Bell was very reluctant to invest time and money in its development. The history of the Airabonita would be an unhappy one.
Vought's Chief designer Rex B. Beisel instead opted for a fuselage of circular cross-section, of a diameter matching that of the R-2800. The oil cooler and supercharger air intakes would be installed in the wing leading edges. He also avoided the hump-backed upper fuselage of the Grumman F4F and F6F, that was designed to give the pilot a better forward visibility over the engine. Hence, the forward fuselage was of cylindrical shape. Construction was all-metal, and streamlining was improved further by using a new spot-welding technique that gave a very smooth finish.
A very large propeller was required to convert the power of the R-2800 into forward thrust. A three-bladed propeller with a diameter of 4.04m was chosen. Sufficient propeller clearance could have been achieved by designing a long and stalky landing gear, or by making the fuselage deeper again, thus moving the wing downwards relative to the engine. Instead, the Vought team adopted an inverted gull wing: the wing started with strong anhedral, i.e. a downwards slope toward the wingtips, and then curved upwards to strong dihedral. The landing gear was installed at the lowest point of the bend. Such a construction was not uncommon, though usually associated with fixed landing gear, such as on the German Junkers Ju 87 "Stuka" dive bomber. Inevitably, the weight of such a construction is higher than that of a straight wing. But apart from keeping the landing gear short and simple, it offered the advantage that the joint between wing and fuselage was made at the ideal angle. In that way a wing root fairing could be avoided. The entire construction contributed to the purposeful ugliness of the design, but it was efficient.
The wing had integral leading edge fuel tanks, which were unprotected. For storage aboard carriers, the wing folded upward outboard of the main landing gear legs. The wheels folded backwards, turning through 90 degrees while retracting, so that they were stored flat within the wing. The entire trailing edge inboard of the ailerons was provided with flaps. The outer wing panels were covered with fabric aft of the wing spar.
The pilot sat in a large cockpit over the wing trailing edge. The view straight forward over the engine cowling was poor, even more so than common in single-seat fighters of the day. View too the sides was reasonable, although the cockpit canopy was heavily framed. No concessions were made to rearward view, the aft of the cockpit being faired into a gentyly sloping fuselage decking. The tailplanes and fins had rounded tips, and the control surfaces were fabric covered.
Armament consisted of one .50 gun in each wing, and a .50 and a .30 in the engine cowl decking. There was also room for 20 small anti-aircraft bombs, stored in the wings.
First flight of the XF4U-1 was made on 29 May 1940, by Lyman A. Bullard Jr. The XF4U-1 was powered by a XR-2800-4 engine, rated at 1805hp. The first flight was not uneventful. A hurried landing was made when the elevator trim tabs failed because of flutter.
Early testing encountered a serious setback when project pilot Boone T. Guyton ran out of fuel during the fifth test flight and made an emergency landing on a golf course. The XF4U-1 was badly damaged, but not beyond repair, and Chance Vought rebuilt it.
On 1 October the XF4U-1 made a flight for Stratford to Hartford with an average ground speed of 650km/h (404mph). It was then the first US fighter to fly faster than 400mph. The XF4U-1 also had an excellent rate of climb. On the other hand, the testing of the XF4U-1 revealed that some of the requirements of the US Navy would have to be rewritten. In full-power dive tests speeds of up to 885km/h were achieved, but not without damage to the control surfaces and access panels, and in one case, an engine failure. The spin recovery standards also had to be relaxed, as recovery from the required ten-turn spin proved impossible without recourse to an anti-spin chute.
Much time was spent trying to improve the handling of the XF4U-1. Numerous changes were made to the ailerons, with success, as these were later known to be very effective. However, the low-speed handling characteristics left much to be desired. The F4U had a troubling tendency to drop a wing when it stalled. And this was a critical factor for a shipboard fighter, which would have to make dangerous deck landings.
For the production F4U, the US Navy required some changes, which were logical in itself but had unfortunate side effects. More armour was carried for the pilot and oil tank, which added 68 kg to the weight. The armament was changed to six .50 machine guns, three in each wing. The wing bomb bays were deleted. This increase in firepower was needed, but the wing guns displaced the leading edge fuel tanks. To restore an adequate fuel capacity, an additional fuel tank had to be installed in the fuselage. Because it had to be near the center of gravity, there was no other option than moving the cockpit to the rear. The 897 liter self-sealing fuel tank pushed to cockpit 0.91 meter closer to the tail. Forward view over the engine cowling, already poor in the prototype, was now decidedly bad. This was especially a problem during take-off and landing, because the F4U, like most fighters of its generation, was a tail-dragger. On the other hand, rearward vision was improved a bit by making cutouts in the rear fuselage decking. Vision to the sides and downwards was excellent.
The ailerons were enlarged, the cockpit canopy was made jettisonable, an IFF transponder was fitted, and the tailwheel design changed. The engine of production aircraft was the R-2800-8, rated for 2000 hp at an rpm of 2700 for take-off. It had a mechanical two-stage, two-speed supercharger. When all changes were incorporated, the gross weight had increased considerably. The XF4U-1 had weighed 4244 kg, but the F4U-1 5758 kg.
The performance of the F4U was impressive. Below is a comparison with the two other fighters which were powered by the R-2800. The F4U was considerably faster than the competing F6F Hellcat. It was slower than the P-47 Thunderbolt, but the latter achieved it highest speed at 9150m, with the help of a turbocharger. The F4U had a mechanically supercharged engine.
|Engine||Pratt & Whitney
|Pratt & Whitney
|Pratt & Whitney
|Wing Span||12.49 m||13.08 m||12.43 m|
|Wing Area||29.17 m2||31.03 m2|
|Length||9.99 m||10.17 m||11.00 m|
|Height||4.58 m||4.40 m||4.35 m|
|Empty||4025 kg||4101 kg||4485 kg|
|Loaded||5388 kg||5528 kg||6116 kg|
|Max||6280 kg||5997 kg||6795 kg|
|Speed||631km/h at 7315m
515km/h at S/L
|605km/h at 6950m
521km/h at S/L
|676km/h at 9150m
|Climb||3050m in 5.1min
6100m in 10.7min
|4575m in 6.0min
7620m in 11.3min
|4575m in 7.2min
6100m in 11min
9150m in 20min
|Ceiling||11310 m||10820 m||12800 m|
|Range||1722k m||1746k m||1344k m|
|Armament||Four .50 M2 with 400 rpg
Two .50 M2 with 375 rpg
Two 454kg bombs
|Six .50 M2 with 400 rpg||Eight .50 M2 with 300 rpg|
Overall handling of the F4U-1 was acceptable, but not very good. In level flight the Corsair was stable enough to be flown hands-off. The ailerons were light and effective, and the high roll rate was used with good effect in combat with the A6M, which suffered from bad aileron response at high speeds. The elevators were heavy, but effective. Only the rudder really stiffened with increasing speed. For combat maneuvering, the flaps could be deployed 20 degrees.
After the first delivery of an F4U-1 on 31 July 1942, more than two years passed before the US Navy cleared the type for shipboard operations. The Corsair was found to be much too difficult to land on a carrier deck. First of all, the pilot could hardly see the deck, because he sat so far aft of the bulky engine. The F4U tended to stall without warning, and was then certain to drop the starboard wing. Quick action had to be taken to prevent a spin. Spin recovery was difficult. In landing configuration, the F4U-1 would stall at 141km/h. A warning light would light at 148km/h. On touchdown, the F4U-1 had sluggish controls and insufficient directional stability. It also was prone to "bounce" because of overly stiff landing gear oleo legs.
These characteristics had already been there on the XF4U-1, and if anything they were worse on the production type. Carrier qualification trials on the escort carrier USS Sangamon Bay, on 25 September 1942, caused the US Navy to release the type to the US Marine Corps. After all, the US Navy still had the Grumman F6F Hellcat, which did not have the performance of the F4U but was a far better deck landing aircraft. The Marines needed a better fighter than the F4F Wildcat. For them it was important that the F4U could be put on a carrier, but they usually flew from land bases.
The islands from which the advanced units operated were often very small. If they were larger, they were often covered with a dense jungle, and only a small part of the island was used by the combattants. The climate was often unhealthy, both for people and aircraft, and standards of living were primitive. Missions often involved long overwater flights. The island group of the Eastern Solomons, for example, extends over more than 1000km.
The first USMC unit to equip with the F4U was VMF-124, which was declared operational on 28 December 1942. VMF-124 was quickly deployed to Guadalcanal, where it flew its first combat mission, also the first of the F4U, on 11 February 1943. Fighting over Guadalcanal was intense. The first air-to-air combat took place on the 14th, when a mixed force of P-38s, P-40s, PB4Ys and F 4Us lost ten aircraft to the Japanese, and claimed four A6M "Zero" fighters.
As on this first mission, the aircraft involved in an operation were often of different types, belonged to different services, and belonged to different bases. The coordination between them was not always what it should have been.
Within six months, all USMC units in the Pacific were equipped with the F4U. The production was extremely rapid, and by August 1934 a thousand aircraft had been delivered. Final production of the F4U-1 was 5559, including the 2010 FG-1s built by Goodyear and 735 F3A-1s built by Brewster.
Fleet Air Arm units where created and equipped in the US, at Quonset Point or Brunswick, and then shipped to war theatres on board of escort carriers. The first Corsair unit of the FAA was No 1830 Sqdn, created on the first of June 1943, and soon operating from HMS Ilustrious. At the end of the war, 19 FAA squadrons operated with the F4U. British Corsairs operated both in Europe and in the Pacific. The first, and also most important European operations were the series of attacks in April, July and August 1944 on the German battleship Tirpitz, for which Corsairs provided top cover. In the Pacific the FAA Corsair also began to operate in April 1944, participating in an attack on Sabang, and later in the attack on oil refineries at Pelambang.
In July and August 1945, the Corsair squadrons No 1834 , No 1836 and No 1842 took part in a series of strikes on the Japanese mainland, near Tokyo. They operated from the carriers MHS Victorious and HMS Formidable.
The other major user of the Corsair was New Zealand. It received over 425 F4U-1A and F4U-1D models. In late 1944 the F4U equipped all twelve Pacific-based fighter units of the RNZAF. The first squadrons to use the Corsair were Nos 20 and 21, on Esperitu Santo island, operational in May 1944. In the RNZAF Corsair units, only the pilots and a small staff belonged to the squadron; aircraft and maintenance crew were grouped in a pool.
The RNZAF Corsair mainly flew close-support missions, and as a consequence did not claim a single enemy aircraft shot down. At the end of 1945, all Corsair squadrons but one (No 14) were disbanded. That last squadron was based in Japan, until the Corsair was retired from service in 1947.
The cowling gills on top of the fuselage were soon fixed in the closed position, to avoid the deposition of oil on the windscreen. For similar reasons, the joints of the fuselage fuel tank in front of the cockpit were often covered with sealing tape.
To cure the tendency to drop a wing, a small spoiler was installed on the starboard wing.
From the 759th aircraft onwards, the framed canopy of the F4U-1 was quickly replaced by a much neater plexiglass "bulb" with small frames. The raising of the seat by 18cm slightly improved the view over the nose, and the new type also offered some rearward vision. Later this modification was associated with a change of designation to F4U-1A, which was not used at the time.
An important change, from the 1550th aircraft, was the installation of the -8W engine with water injection, which allowed higher emergency power to be used at low altitude.
The F4U-1B designation seems to have been used for the F4U-1As delivered to Britain.
The F4U-1C had four 20mm cannon instead of the six .50s. These guns were the British Hispano Mk.II cannon, known in the USA as the Hispano M2. These weapons protuded far from the leading edge. Production of this version remained limited to 200. They entered combat in April 1945.
In early 1944, bomb racks for the F4U-1 were developed by personnel of VMF-222 and VF-17. The modification was rapidly applied by other squadrons. The F4U-1D was a factory-built fighter-bomber model, powered by a R-2800-8W engine with water injection. The F4U-1D had three pylons, one on the centreline and two on the wings. Later small stubs on the outer wing panels, to carry rockets, were added.
Also in early 1944, longer oleos were installed in the main landing gear legs. They cured much of the tendency of the Corsair to "bounce". A longer tailwheel leg raised the fin, and reduced the directional stability problem.
These improvements were essential in making the Corsair suitable for carrier operations, and in April 1944 the Corsair was finally qualified for carrier operations.
An F4U-1 with a special mount in the rear fuselage for a K-21 camera was known as F4U-1P. The F4U-1P was used mainly to assess the results of air strikes. No F4U-1Ps were produced by the factories, they were all modified in the field by USMC or USN units.
In November 1943 the land-based VF-17 ran out of fuel while giving top cover to the carriers USS Essex and USS Bunker Hill. The aircraft then landed on the carriers, without incidents.
The US Navy finally accepted the F4U for shipboard operations in April 1944, after the longer oleo leg was fitted, which finally eliminated the tendency to bounce. The first Corsair unit to be based effectively on a carrier was the pioneer USMC squadron, VMF-124, which joined the USS Essex. They were accompanied by VMF-213. The increasing need for fighters, as a protection against Kamikaze attacks, resulted in more Corsair units being moved to the carriers.
The Navy squadrons VF-12, VF-17 and VF-301 also soon operated from carriers.
The original radar was the AIA installation, developed from the Massachusetts Institute of Technology under Project Roger. It had a range of 6km against aircraft. For single-seat fighters a easy-to-use scope had to be developed. This took the form of a small circular scope on the instrument panel that showed two blips for the target. The first blip indicated the direction and distance of the target, and the position of the second blip relative to the first one was an indication of the relative height of the target.
The small radar radome was added on the starboard wing, on the wing leading edge close to the wing tip. To compensate for the weight one of the wing guns was removed, and ammunition reduced. As the F4U-2 was intended for night operations, flame dampers were fitted to the exhaust stacks. A radio altimeter and an autopilot were also installed.
The F4U-2 equipped VMF(N)-532, VF(N)-75 and VF(N)-101. Early operations of VF(N)-75 in New Georgia revealed considerable problems with the operating procedures, but on the night of 1 November Lt. O'Neill shot down a G4M bomber. The tactics finally developed let the F4U-2 climb towards its target from astern. This also helped to decelerate the fighter enough, to prevent it from overshooting its target.
VF(N)-101 was created by splitting of part of VF(N)-75. It was the first carrier-based nightfighter unit of the USN. This was in January 1944, and made the unit the first carrier-based Corsair squadron. A limited number of night operations was flown, because of reluctance to take the risk. Nevertheless, no accidents occured, which helped to clear the Corsair for carrier operations.
Nevertheless, the Navy preferred to develop a nightfighter version of the F6F Hellcat, which was easier to fly and to deck-land. For night operations those were important advantages, and the Hellcat became the standard single-seat nightfighter.
After the three XF4U-3s, only a single Goodyear-built FG-1A was converted to FG-3, before the programme was cancelled. Twelve more FG-3s were completed, but were used only for development work.
The early R-4360-4 engine was rated at 3000hp. Because of the greater length of the four-row R-4360 radial, the engine cowling of the F2G was elongated. Together with the air intakes behind the engine cowling, on top of the fuselage, this was an easy recognition feature. The tail surfaces were enlarged, and more fuel capacity wa installed. Goodyear also fitted an all-round vision bubble cockpit on the F2G. This had first been tried on a FG-1A. It was a significant improvement, that for some reason was not adapted by later models of the Corsair.
The first models were land-based F2G-1s, but they were later followed by F2G-2 carrier fighters with hydraulic wing folding.
Production of the F2G ended after eight prototypes, five F2G-1s and five F2G-2s were completed. The original order for 418 F2G-1s was cancelled, because the end of the war removed any need for the F2G. Climb was excellent, 9150m could be reached in 4 minutes. Maximum speed on the other hand was rather disappointing, 32km/h (20mph) down from the expected 724km/h (450mph). The F4U-5, with its uprated R-2800 engine, was faster than the F2G. The F2G also suffered from lateral control problems.
During the F4U-4 production, the cockpit was redesigned again. It now incorporated a flat, bullet-proof windscreen, a revised canopy, an armoured seat, and an improved instrument panel.
Production included 2050 F4U-4s with six .50 guns, 297 F4U-4Bs or F4U-4Cs with four 20mm cannon, a single F4U-4N nightfighter conversion and nine F4U- 4P reconnaissance modifications. The last one was delivered in August 1947. Plans to produce the F4U-4 by Goodyear as the FG-4 were abandoned.
The F4U-4 arrived late in WWII, and served only during the last four months of the conflict. The war of the F4U-4 was the Korean war. Here the type served mainly as a fighter-bomber, but nevertheless one pilot, Capt. J. Folmar of VMA-312, was credited with shooting down a MiG-15.
The first XF4U-5 flew on 4 April 1946.
There was also a nightfighter version, the F4U-5N. The radar was again, as in the F4U-2, installed on the outer starboard wing. The radome was different in shape, however, betraying the presence of the improved AN/APS-6 and later AN/APS-19A radar set. The AN/APS-6 radar had a range of 8km against aircraft, and 37km against ships.
Production included 223 F4U-5s, 214 F4U-5Ns, and 30 F4U-5P reconnaissance models. In addition 101 winterized F4U-5NLs were built, with de-icing booths for service in the bitter winters of Korea. Production continued until October 1951.
In the late 1950s the US delivered a small number of F4U-5s and F4U-5Ns to the Argentine Navy.
The AU-1 was given more armour for the pilot and the engine. Four 20mm cannon with 231 rounds each were installed in the wings. The number of outer wing racks was increased from eight to ten.
Performance had, of course, decreased. The handling had suffered even more, and the AU-1 was unpleasant to fly. Only 111 were built between February and October 1952.
In addition, the French acquried a few AU-1s used previously by the USMC.
French Corsairs fought in Indochina, Algeria, and the Suez conflict. The last were retired in 1964.
The F4U is often said to have been the most successful fighter of WWII. This is based on a claimed 11 to 1 kill ratio: 2140 enemy aircraft shot down for a loss of 189. But as a measure of effectiveness, this is not very reliable. Kill claims are almost invariably too high: Repeated firing on the same aircraft, the confusion of a fast-moving battle, overestimation of damage done to the enemy, and over-confidence of the pilots usually produce estimates which are at least a factor two too high. Also, the opponents encountered by the Corsair squadrons in the Pacific were of greatly varying quality. Most of them indeed flew aircraft to the F4U, but the A6M "Zeke" was inferior to all US fighters of the end of WWII. Finally, to these 189 lost in air-to-air combat one should add the 349 shot down by anti-aircraft fire, the 164 that crashed on landing, and the 992 that were lost for other reasons, including training accidents. The large number of aircraft lost to anti-aircraft fire reflects the use of the Corsair as a fighter-bomber. In this role it excelled, and its use continued into the Korean war.
In one respect the F4U must be considered a partial failure: More than two years passed before the Corsair became an acceptable deck-landing aircraft. Of the 64051 combat missions flown by the type in World War II, 54470 were flown from land bases. If Grumman had not hastily produced the F6F Hellcat, the US Navy could have been in serious trouble. Clearly the Hellcat and Corsair represented different design philosophies: The Hellcat sacrificed performance to simplify production and to make it a better deck-landing aircraft, but the Corsair did not. The Grumman team also produced a fighter that was almost right from the start, apart from the engine change in the early stages of development. Far more time and effort were required to realize the potential of the Vought fighter, and by the time it was fully developed the war was almost over.
Production of the Corsair ended after 12571 had been built, which 4017 by Goodyear and 735 by Brewster. For comparison: Grumman built 12275 Hellcats, and Republic completed 15683 Thunderbolts. One must take into account that production of the F4U continued after the war, and that of the F6F and P-47 did not. The actual production rate of Vought was lower than that of its competitors, but it was still impressive.
|Engine||Pratt & Whitney
|Pratt & Whitney
|Pratt & Whitney
|Speed||631km/h at 7315m
515km/h at S/L
|756km/h at 8170m
649km/h at S/L
|694km/h at 5000m
642km/h at S/L
|Climb||3050m in 5.1min
6100m in 10.7min
|?||9150m in 4min|
|Armament||Four .50 M2 with 400 rpg
Two .50 M2 with 375rpg
Two 454kg bombs
|Four 20mm cannon||?|
Text : Emmanuel Gustin