Almost all the first batches of Mk I Beaufighters were night fighters. The earliest deliveries had no radar, but in September 1940 one of them was flown to the FIU for trials with one of the first AI.IV installations. By November 1940 AI.IV was standard on production Mk IF aircraft. The installation was good, and a skilled operator could obtain a clear and unambiguous target blip over a range of at least three miles (provided that the aircraft was about three miles above the ground) down to a minimum range of something like 100 feet. This minimum range was closer than with AI Mk III, but it was still too far for certain visual contact in adverse conditions. One of the most important and most difficult skills an AI operator had to acquire was to judge the rate of closure on the target simply by watching the blips. With the Beaufighter there was ample in-flight performance for overtaking any Luftwaffe bomber of 1940, even if the bomber had become frightened and tried to get away. But it was far from simple to guide the pilot with absolute precision, so that he came up astern to visual contact without overshooting or colliding with the bomber from behind. At that time, sudden closing of the throttles on British engines generally resulted in severe backfiring and large gouts of flame from the exhausts. The Hercules was better than most engines in this respect, but still could not equal the German engines, whose direct-injection fuel systems did not suffer from this failing, which at night could mean life or death. The fact that a night fighter needed powerful airbrakes was not then realized, and much later many Beau pilots learned that the best procedure was to keep on plenty of power and drop the landing gear. The Beau’s great strength is reflected in the fact that the gear-down limit was 240 mph IAS, well above any likely indicated speed at around 15,000 feet (the later Mosquito was red-lined at only 180).
It was thus fairly safe for the pilot to go on closing on the target after the AI blip had vanished off the bottom end of the range scale. With practice one could judge just when the range ought to be dropping close to zero, and visual contact by this time ought to be certain. Should the enemy suddenly appear dead in front, a cool pilot could then slow down by lowering his wheels briefly (though this was not realized during the night Blitz). Once in a firing position the pilot, having made certain that his firing-button outer ring was set from ‘Safe’ to ‘Fire’, could open up with devastating armament. The first fifty Beaufighters had four 20 mm Hispano cannon under the nose, ahead of the hatch. Subsequent deliveries added six Brownings in the wings, four on the right and two on the left.
This armament was the heaviest of any RAF fighter in the Second World War, and still sounds impressive today. In the RAF’s hour of desperate need in 1940 it was manna from Heaven to a service that had nothing more powerful than rifle-calibre machine-guns, which were swiftly becoming ineffective as the Germans bolted armour on to their bombers. What amazed the author, even at the time, was that there was no modern British shell-firing cannon. The 20 mm Hispano-Suiza had been designed at the end of First World War. Installed singly between the cylinder blocks of the same company’s fighter engines, it did not matter that it was eight feet long and weighed 109 lb. Also fitting neatly between the cylinder blocks was the 60-round drum of ammunition. Lacking any obvious alternative, a licence for this was obtained at the proverbial ‘eleventh hour’ in 1937, and by 1939 it was coming into production by British MARC Ltd at Grantham. By 1945 a group of British factories had delivered just under 100,000, later versions having a shorter barrel. Fortunately, the British armament experts showed little interest in the mounting on an engine (and to bolt it on a Merlin would have meant major redesign of the reduction gear). Instead, its first British application was to mount four in the nose of the Westland Whirlwind, though this odd fighter was never of any importance. More to the point, one was – with difficulty, turning the gun on its side – mounted in each wing of the Spitfire IB. The first trial installation took place soon after the outbreak of war, and in March 1940 one of the experimental Mk IBs shot down a Do 17 with sixteen rounds. In the Battle of Britain the Mk IB equipped 19 Squadron, though maddening stoppages and other faults persisted. These were gradually cured, and by 1941 the standard Spitfire fighters were of the B type, with two Hispanos and four Brownings (though in the week the prototype Spitfire first flew, in March 1936, designer Mitchell had schemed a version with four Hispanos!).
Wisely, Bristol could see the futility of the four Brownings of the Blenheim fighters, and replaced them in the Beaufighter by four Hispanos. In this aircraft there was no need for the clumsy drum feed, and the Bristol design team devised a superb installation with continuous belt feed from four 240-round magazines. Air Ministry armament experts rejected the scheme, claiming that it would either jam the guns or be wrecked. Bristol then proposed an air-driven servo feed, tested on the fourth Beau in competition with the established French feed using 60-round drums. In French single-seat fighters only one such drum could be fitted, but in the Beau the observer could leave his radar, clamber forward and exchange used drums for fresh ones. As each drum weighed about 100 lb, and the fighter might be pulling g in a tight turn, changing drums was not popular with the observer (the drum racks had sharp edges, too). More time was wasted comparing two further servo-feeds, by Avro and Hydran. Then suddenly, as France fell, two Armée de l’Air officers arrived with drawings of their own newly developed Chatellerault belt feed. It was at once adopted, though it took until September 1941 to get it into production and in service (on the 401st aircraft). So all the Beaufighters in the long winter Blitz had hand-changed 60-round drums. It was only when Bristol studied the French feed that it was realized it was just like their own, original, rejected feed, except that it extracted rounds from the belt by pushing the nose instead of pulling the case. This proved to be a disadvantage, and it was changed to case-pulling, making it identical to the British scheme that could have been on the very first production aircraft!
Beaufighters became operational during September 1940, and during the winter their numbers increased rapidly, reaching 100 that year and 200 in May 1941. Production of AI Mk IV more than kept pace, and the improved radar was retrofitted to most of the early Mk I aircraft that had been delivered without it. A few Mk IV sets were also installed in Blenheims of FIU, and it was one of these, piloted by none other than Ashfield, that scored the second AI night victory on 7 November 1940. Was Ashfield – who was later killed in action – a superman? At first glance, for the whole RAF night-fighter force to have no success at all using AI, while one crew engaged in research scored two victories, does seem to require explanation. There is always an element of luck in being in the right place at the right time, but I think it is fair to describe night fighting as something uniquely difficult. Until it was mastered, neither a GCI controller nor an aircrew could ever say it was in control of the situation. Then, once one or two exceptional crews had begun to work with gifted controllers, the whole thing snowballed.
In November 1940 the Luftwaffe switched its assault from London to Midlands cities. At that time the RAF was just beginning to discover that its night-bomber crews – thanks to failure to practise before the war – were hardly ever able to find their targets. In contrast, the Luftwaffe had foolproof and easily used radio navaids that had been developed during the late 1930s, based on technology related to the Lorenz radio landing aid. Using X-Gerät and Knickebein, a jumpy and unskilled crew could fly direct to any British city and automatically release their ordnance at the correct point in space for a near-direct hit. These precision radio beams had been reported in the notorious ‘Oslo parcel’. British experts foolishly scorned the notion of the Luftwaffe having in service an advanced aid that had not even been thought of by the RAF, but, fortunately, Tizard sent for young Dr R.V. Jones, the infra-red researcher from the Clarendon mentioned previously, and his methodical sleuthing uncovered the unpalatable truth. Not only did this prove that the German radio technology had been seriously underrated – X-Gerät manufacturers’ date-stamps mostly went back to 1938 – but it enabled Britain urgently to set up jammers and even spoof beacons, which caused more than one Heinkel to come to grief. Unfortunately there was one occasion, on 4 November 1940, when for various reasons no jammers were operating. On that night the city of Coventry was devastated.
Riding their fine Beaufighters, the handful of RAF night-fighter crews who had been converted to the type watched the burning cities, and helplessly chased one ‘contact’ after another (they all just seemed to disappear). Even if they had known that the first waves of bombers were following invisible lines in the sky, like transport aircraft flying the American radio range, they could not have accomplished the vital final step of drawing in close enough behind to see the enemy, identify it and shoot it down. Then, on 19 November, one crew did it. It was 604’s John Cunningham and his observer Sergeant J. Phillipson: during a raid on Birmingham they shot down a Ju 88. Chisholm, also of 604 Squadron, wrote, ‘The news was electrifying. For me it meant that the bombers we were sent to chase were really there . . .’ In the minds of the NF crews, the Magic Mirrors, with their overtones of useless trickery, had become AI Mk IV, a system which, placed in the right position by increasingly adept ground control, could – with a modicum of luck – take a good crew to a visual contact. Cunningham quickly gained other successes, became a national hero, and for the rest of his life suffered being called ‘Cat’s Eyes’. Bombers were being shot down at last. With radar unmentionable, a common Ministry of Information cover story at the time was that night-fighter pilots not only wore dark glasses before take-off, to become night adapted (which they certainly did), but also derived abnormal night vision by eating copious quantities of carrots (which most of them did not). Even today there are thousands of the British public who have a lingering image of cat-eyed pilots munching this vegetable.
A vital role in closing the last links in the world’s first night defence system worthy of the name was provided by the GCI stations previously mentioned. It was essential for the ground controller to have an up-to-date (real time) picture of the local situation; a lag of thirty seconds could well make interception impossible. What had been AMRE at Bawdsey, and was now renamed TRE (Telecommunications Research Establishment, a pure ‘cover’) at Worth Matravers, played the chief role in creating the GCI radar, with its PPI (plan-position indicator) picture painted by successive sweeps of a rotating radar beam, just as in so many radars today. Some of the GCI team were in the wooden huts on the Dorset coast, and others at the airfield at Christchurch – rather an inconvenient distance on the other side of Bournemouth – where TFU (Telecommunications Flying Unit) kept its growing fleet of night fighters and other trials aircraft. The PPI was swiftly improved until each sweep took sixty-five seconds, instead of the original eight minutes, and reached out to a distance of ninety miles. It could not indicate height, which had to be telephoned in by a CH station.
TRE undertook a crash programme to make the first dozen GCI sets, often using any suitable hardware that was handy, and delivered the first on 16 October 1940, at Durrington, Sussex, where it was operational two days later. The last of this batch was operational by 6 January 1941. Percival Rowe, by then Superintendent of TRE, recalled:
In theory, it looked simple enough to us. All that was needed was for a night fighter to patrol up and down the English Channel until told by R/T from the Worth Matravers GCI station that there was an enemy bomber in a suitable position for interception. Bomber and fighter would then be tracked and their heights assessed. When the fighter had been put on the tail of the bomber by directions from the GCI station, the latter would give the signal ‘Flash weapon’, which would tell the fighter crew that the AI set should be switched on and used to complete the interception. The time was to come when it really was almost as simple as this, but not at Worth Matravers.
The civilian boffins at that establishment were the very first controllers to direct night fighters using the original prototype GCI, but there seemed to be too many problems for a kill to be achieved. Not least of the problems was that, like Chisholm near his home airfield, the night fighters continually got lost. There were three important answers. The first was a regular pattern of bonfires across southern England maintained by the ROC (often at peril to their lives from people who thought the blackout was being sabotaged). The second was the building of the Racons – passive ground radar beacons – which showed up strongly on the fighter’s own AI scopes. The beacons were arranged in a regular checkerboard pattern across southern England, and the fighter observer could always see the blip from at least one and usually from two. By measuring the distances it was possible to get a fair idea of position, refined if necessary by actually plotting with a compass and pencil on a topographic map (not a plotting chart). The third way of solving the navigation problem was to learn to trust the GCI controller who, once the PPI sets were in use, always knew where the fighter was; he could identify it by its IFF signal. If required – and when lost on a dark night one is less concerned about loss of face – the controller could talk the fighter pilot home to his own airfield. To quote Rowe, ‘Once this had been demonstrated, the fighter pilots poured over from Middle Wallop to see the new brand of magic, after which they came to believe that, though they might not know where they were, there was an “eye” on the ground that could watch them and lead them home.’ This again was a fundamental advance in the technology of aviation that today is taken for granted.
This ability to be guided home was especially helpful to the pilots of single-seaters and Defiants, which had nobody on board to help the pilot navigate. With Racons and GCI radar the concept of the AI-equipped single-engined fighter was looked at again, and it was soon realized that AI could be put in the Defiant and the Hurricane after all. Boulton Paul Aircraft were given drawings of the pilot-indication AI Mk V, which had a novel CRT scope with a U-marker around the blip, and the company completed installation drawings for the Defiant on 19 November 1940. The aerials were as in the single-seat scheme of the previous July, with the transmitter on the right wing. The transmitter itself was behind the turret, the receiver, control box and power pack behind the pilot’s seat, the display on the pilot’s left, and the control panel on his right. This pioneer single-engine installation suffered from moisture and bad electrical screening, and for some reason was not cleared for service until August 1941. By this time the radar had become AI.VI, with a wider bandwidth of 188–198 MHz and an added beacon facility. The radar-equipped Defiant IA served with 264 Squadron and later with 96, 125, 256 and 410. The more powerful Defiant Mk II served additionally with 141, which had destroyed an He 111 the previous December without AI, and with 151 and 153 Squadrons.
Having discovered that there was no law of nature forbidding radar in a single-engined aircraft, the Air Staff also fitted AI.VI to twelve Hurricane Is in late 1941. In November 1942 a report on their first year of operational service simply glowed with enthusiasm, especially commenting on their high serviceability and rapid accumulation of flight time, leaving the impartial observer mystified at why it was not done before and not done again.