This map shows the German nightfighter defences in early 1942. The coastal chain of dark fighting Freya-AN Dunaja zones is backed by a line of Himmelbett boxes ranging from Denmark into France. Each Freya and Himmelbett station could control only one fighter at a time. By mid-1942 the searchlights had been withdrawn to the cities, creating large illuminated zones (here marked in a lighter green) where Konaja, and later Wilde Sau, fighting could take place.
On 14 May 1940 the Luftwaffe set out to break the dogged Dutch hold on the north bank of the Maas by a mass attack on Rotterdam. As the bombers neared their target, the Dutch opened negotiations for surrender. The Luftwaffe recalled its bombers, but one unit – the fifty-seven Heinkels of KG 54 – had already done its work and started fires which gutted the heart of the Old City. Far worse attacks had been made on Polish cities, and later on Belgrade, but for some reason this incident made the Allied leaders recoil in horror. On the following day Winston Churchill, just appointed British Prime Minister, announced that henceforth the RAF could bomb Germany. That night ninety-six heavies set out against specific oil and rail targets in the Ruhr. Only twenty-four even claimed to have located their objectives.
Later, a bomber pilot visiting TRE said, ‘They used to tell us to bomb Krupps, but we were lost as soon as we left the aerodrome.’ Clearly, Bomber Command had a long and difficult road ahead; but so did the Luftwaffe. Reichsmarschall Hermann Goering, the famed head of the Luftwaffe, had promised in August 1939, ‘We will not expose the Ruhr to a single bomb dropped by enemy aircraft.’ He had just inspected some of the Luftwaffe’s heavy Flak emplacements near Essen, with 88, 105 and 128 mm guns radar-directed by the new Würzburgs. By a very wide margin indeed, it was the best AA artillery in the world. But radar-directed Flak took a long time to come into general use, and the Luftwaffe had no night-fighter aircraft at all. In its first four months of unrestricted bombing of Germany between May and mid-September 1940 the RAF lost only 163 aircraft, about two per cent of the 8,000-odd sorties. Goering was bothered, because the occasional bomb was falling on the Ruhr; one or two even hit their intended targets. In July 1940 he instructed Colonel Josef Kammhuber to form a special force of night fighters.
Kammhuber was not impressed by the existing German defence system. The Würzburgs and heavy Flak formed a formidable combination, but there were still only 450 guns and a mere handful of radars. Night fighters were another story. Nobody in the Luftwaffe had even dreamed of putting radar into a fighter, and the only method of operation was Helle Nachtjagd (illuminated night fighting). A few day fighters, nearly all Bf 109s flown by bolder or more experienced pilots, would take off on radar early-warning of a British raid and orbit a radio beacon. Often they would keep their navigation lights on to avoid a mid-air collision, and sometimes they would have to fly to a second beacon and orbit again. Interceptions were achieved solely by watching the searchlights and trying to see the enemy bombers. Throughout the summer this method resulted in just one success: on 9 July 1940 Feldwebel Foerster of JG 2 managed to shoot down a Whitley. He would probably have admitted that this was mainly by luck.
Kammhuber could see the need for bigger twin-engined night fighters, with adequate endurance for the long night patrols. The Bf 110 was an obvious choice, but an even better one might be the Ju 88C. This sub-type had begun life as a long-range day and anti-shipping fighter, with the Ju 88V7 prototype flown on 27 September 1938, which had an unglazed nose mounting two cannon and two machine-guns. Some pre-production C-0 fighters were used for ground attack during the Polish campaign, but plans to build the fast C-1, with two BMW 801 radials, were shelved to enable all effort to be applied to building A-series bomber versions. In 1940 the BMW 801 went into production, but with priority for the Fw 190 single-seater, so the C-1 was abandoned and instead the Luftwaffe began to receive the Ju 88C-2, a rather hasty conversion of the A-1 bomber with Jumo 211 engines and a nose armament of one cannon and three machine-guns. Instead of being ordinary day Zerstörer (destroyer) fighters, these were now regarded as primarily for use by night. They were the forerunners of the aircraft that were to play the biggest part in the biggest night air battle in history.
While giving much thought to the last long-term system for the night defence of the German-controlled continent, Kammhuber acted quickly to create a night-fighter force. On the day of his appointment he picked the premier Bf 110 Staffel, I/ZG 1 commanded by Major Wolfgang Falck, and transferred it to Düsseldorf to serve as the nucleus of a night-fighter research and training school, with the unit designation of NVS 1 (Nacht und Versuchs Staffel). Three days later it was redesignated I/NJG 1 (Nacht jagdgeschwader = night fighter wing), on 20 July 1940. Falck was promoted Geschwaderkommodore of NJG 1, and a second squadron, II/NJG 1, was formed with twenty newly delivered Ju 88C-2s. Hauptmann Gunther Radusch took over I/NJG 1, and the force swiftly expanded by adding III/NJG 1 with Bf 110Cs from IV(N)/JG 2, and IV/NJG 1 from Zerst Sta/KG 30 with Bf 110Ds; a fifth staffel was also added, partly based on a special unit that had been formed to operate the first Dornier night-fighter conversions, the Do 17Z-6 Kauz 1 (Screech Owl I) and Do 215B-5. On 11 September II/NJG 1 was redesignated as the nucleus of the second wing, I/NJG 2, and a new II/NJG 1 was promptly formed from I/ZG 76, one of the most famous Bf 110 units (they shot down the twelve Wellingtons in December 1939) under Hauptman Graf von Stillfried. Another unit became the nucleus of a third wing, I/NJG 3, with Radusch taking over as Kommodore.
Kammhuber was eventually promoted Major-General and set up his HQ in the beautiful castle at Zeist in Holland. He reported to Colonel-General Hubert Weise, in overall command of the German air-defence organization. With constant changes and improvements, the main effect in the first few months, in the autumn of 1940, was the rapid build-up of a well-equipped force of large night fighters, each with a pilot and observer (and gunner, in the case of the Ju 88s and Dorniers), heavy nose armament and endurance of seven hours. In general, the Bf 110 units were deployed geographically to intercept bombers already over Germany, and were alerted by Freya early warning and guided to their targets by the Freyas and the searchlights. The bigger Ju 88 and Dornier aircraft operated around the periphery of Europe in the intruder role, unhesitatingly following bombers right back to their English bases if necessary, and using bombs as well as guns. At this time the Luftwaffe was indisputably the supreme air force in the world. It was easily the best equipped, and the unpalatable failure to subdue the RAF by day had not noticeably affected its morale. It had an abundance of skilled crews, and it was still conditioned to believe in a succession of swift victories (one of its few shortcomings was that no provision had been made for a long war). Not least, it had the backing of a large and competent equipment and radio industry, and by a rapidly increasing margin the world’s best aircraft guns.
What was less good was the makeshift night interception system. None of the fighters yet carried their own radar, so they were strongly dependent upon the searchlights. The latter were grouped around the target cities, so not much could be done to intercept the bombers on their flight to and from their targets except as they crossed the belt of searchlights along the coastline. Over the target the sky was full of Flak, and at that time there was no way for the German Flak to tell which were RAF bombers and which were NJG fighters. Accordingly, during September 1940 Kammhuber took the bold decision to move nearly all his searchlights from the cities to a single dense belt stretching from Liège (Belgium) to Schleswig-Holstein (near Denmark). Virtually all RAF bombers had to pass through this belt, within which no German aircraft were permitted after dark except NJG fighters on patrol. This immediately stopped the wastage of night fighters shot down by their own Flak, but it was by no means a complete solution. The Flak gunners now had hardly any searchlights, and were still waiting for their Würzburg radars. And it needed only a thin cloud layer to wreck the whole system.
It was obvious that what was required was a more sophisticated defence using Würzburgs not only to direct the Flak but also to direct individual night fighters. This radar sent out a fine pencil beam focused by a large circular dish reflector. Nothing like it had been seen before, and as the movable dishes gradually appeared all over northern Europe they excited much comment, most of it concerned with ‘giant mirrors’. The bearing and elevation of the aerial could be read off with great accuracy, and the discrimination was good enough to distinguish two aircraft less than 500 feet apart at over 20,000 feet. On the other hand Würzburg’s extreme limit of range of 25 miles meant that Freya would be needed to give early warning, and get the Würzburg and night fighter into the right positions beforehand. Perhaps the biggest problem was the inability of Freya to indicate the hostile target’s altitude. The night fighter would therefore have to scramble and climb up to a likely altitude by guesswork. In 1940 a good attacking height for a Wellington or Hampden was 15,000 feet, with a Whitley appreciably lower. Only in the final few minutes could the Würzburg suddenly pass an accurate height.
In September 1940 the first trials took place using night fighters directed by a ground controller. Luftwaffe fighter pilots argued heatedly about the supposed loss of initiative and freedom of action in accepting such control – a psychological problem that was much less evident in Britain at this time – and the record shows that the Germans were at first far from eager to accept any of the new radar methods. The first GCI radar tried by the Luftwaffe was a Freya, excellent for early warning but hopeless in the GCI role, because no controller could separate the fighter’s blip from that of the bomber once the range had closed within a mile. Despite this, it was a Freya that was rigged up near Zwolle, Holland, together with a naval height-finding radar, and trials began against ‘Auntie Ju’ (Ju 52/3m) transports. On the whole they were as unsuccessful as those in Britain at this time, but on 16 October 1940 Leutnant Ludwig Becker of IV/NJG 1 suddenly found himself in visual contact with an unidentified aircraft flying east over Holland. Flying a Do 215B from Gilze-Rijen, Becker closed slowly and identified the aircraft as a Wellington. With little difficulty he hit it hard in a five- or six-second burst and watched the bomber eventually spin into the ground. But this was the exception that proved the rule, and it was gained in bright moonlight.
By 1941 Kammhuber had masterminded a completely new defence system, and his organization had placed large orders for an improved GCI radar, Gigant (giant) Würzburg. The need for such a radar was obvious, because Freya had inadequate accuracy and discrimination, and Würzburg had inadequate range. It was not uncommon for RAF bombers to pass through the defence belt while the NJG fighters were still trying to reach the same approximate position and height. A further problem with Würzburg was that reflections from the ground began to mask the target blip at flight levels lower than 6,000 feet (though, of course, few RAF night attacks came down as low as this). Gigant Würzburg accordingly had a much larger aerial dish, roughly twenty-five feet in diameter compared with Würzburg’s ten feet, which concentrated the energy into a narrower beam capable of giving a clear blip at a range of more than forty miles with typical aircraft targets. Telefunken hurried the improved set into production at the end of 1941, by which time the Luftwaffe had placed large orders.
Kammhuber needed several hundred Gigant Würzburgs to equip his grand design to defend the Reich, which became popularly known to the RAF as the Kammhuber Line but was officially designated Himmelbett (heavenly bed, i.e., a four-poster). This code-name stemmed from the fact that Kammhuber divided up the airspace round the north and west sides of Germany into notional boxes, each having a rectilinear shape like an old four-poster. Each box was about twenty miles wide, and there were 750 of them strung in a vast curve from Denmark round the north of Germany, across the Low Countries and south through eastern France to Switzerland. Somewhere in each box was a GCI station equipped with a Freya early-warning radar, a Gigant Würzburg to track a chosen bomber, and a second Gigant Würzburg to track the NJG fighter assigned to that box. At least, that was the intention; the tracking radars were mainly earlier Würzburgs until well into 1942.
Himmelbett had many good features. First, each box was a definite functioning system, technically capable of putting a night fighter very accurately onto the tail of a hostile bomber. It had enough width, something like 150 miles of electronically guarded sky, for there to be plenty of time to set up the interception long before the raider had passed out of the box, let alone out of radar range. And as Kammhuber set up the line just outside his searchlights, the latter were ready to take care of any bombers that the night fighters missed under GCI. By this time the searchlights were arranged in groups of five, one of which was a radar-directed master. The latter, with a brilliant beam having a bluish tinge, was alight all the time, normally pointing straight upwards. Once the associated Würzburg had locked-on to the bomber, the master searchlight would suddenly swing right onto it, much too fast to be dodged. At once the other four beams would snap on and light up the unfortunate bomber; then the master would return to the vertical, waiting for the next customer. Whether this was done to aid fighters or Flak, it was heartily disliked by the Bomber Command crews. Only an exceptional pilot could shake off the cone of beams on a clear night, and it made night-fighter interception almost easy.
On the other hand, there were plenty of shortcomings in the system. It could handle only one bomber at a time per box, and the most rapid interception rate a skilled set of Himmelbett and night-fighter crews could possibly hope for was six aircraft per hour for any single box. In 1941 this was not a serious problem, because the lumbering twin-engined heavies crossed enemy territory at about 165 mph in a thin stream, often many miles apart, generally unsure of their position, and often having to spend as long as an hour taking astro shots, working out revised winds and searching for their target. One crew in a Whitley actually spent over 2¾ hours in the general target area trying to find the place they had been sent to bomb (München-Gladbach). In its first eighteen months the Kammhuber Line was able to pay close attention to the majority of the RAF bombers that attempted to cross it, but the situation was to change dramatically.
A less apparent drawback was that, almost unbelievably, the Gigant Würzburg perpetuated a basic feature of the earlier radar which made it unsuitable for the GCI function. The original Würzburg had been designed for directing Flak, and accordingly gave its information in the form of numerical bearings and ranges. This could easily have been converted in Gigant Würzburg into the ideal form of presentation, the PPI, such as was being used in Britain. Such a display had been developed between 1936 and 1939 by Baron Manfred von Ardenne in his laboratory at Lichterfelde (Berlin) under the name Panorama Sicht Gerät (panorama display equipment). By 1940 he and the short-wave expert Dr Hollmann had prepared this for production with the Radio-Loewe company, and at Christmas in that year a deputation made a presentation to Goering. They explained it in such simple terms that even Goering – the epitome of the technology illiterate, whose opinion of radar was that, ‘It consists of boxes with coils . . . I do not like boxes with coils’ – could not fail to see the advantages. With PPI a controller has a perfect real-time picture, with the aid of which he can use his judgement to tell the night-fighter pilot exactly when to turn, onto what heading, and at what rate, to bring him up astern of the bomber. Goering gradually saw how it worked and what it did, and even he was forced to admit that it was better than a mere list of ranges and bearings. But it was Christmas 1940, and he told the electronics expert, ‘Such a comprehensive development is no longer worth while; the war is already as good as won!’ So Gigant Würzburg provided nothing but ranges and bearings. To provide a PPI picture for the controller, a clumsy device called a Seeburg table was necessary. An operator was told the ranges and bearings of the bomber by telephone, set them up on a rotary and sliding scale in front of him and, in doing so, moved a red spotlight on a large ground-glass table at an upper level. A second operator, connected by telephone to the radar tracking the fighter, moved a spot of blue light in the same way. At the upper level, a third operator with red and blue wax crayons marked the tracks of the two aircraft. The mind boggles at the number of places where errors and inaccuracies could be introduced.
Despite this, the Himmelbett system worked. By the end of March 1942 Kammhuber had about half his initial order for 185 Gigant Würzburgs, and Telefunken was delivering thirty a month. But by this time, in a single bold stroke, the British had made up for their amazingly inept radar intelligence, and learned all they needed to know about the original Würzburg. Though a slight digression from night fighters, it is a thrilling story. For years the British learned nothing about German radar, despite the valuable clues in the Oslo Report. In February 1941 a low-flying reconnaissance Spitfire brought back pictures of circular objects at Auderville, west of Cherbourg, and an interpreter noticed that a narrow object in one of the circles had changed its bearing between one exposure and the next. The British had at last discovered Freya. In November 1941, when a number of Freyas had been pinpointed, interpreters became interested in a small black blob on a path trodden between the cliffs and a large house, at another Freya station at Bruneval, north of Le Havre. Flight Lieutenant Tony Hill went and took pictures at low level with his Spitfire (twice, because the cameras failed the first time) and also had a good look himself. The upshot was one of the earliest and most successful Commando raids ever mounted. On 27 February 1942 twelve Whitleys dropped 119 paratroopers near Bruneval. Next day the 111 survivors, seven of them injured, landed back in Britain with all the vital parts of the Würzburg, plus three prisoners, one of whom was a skilled radar operator. In subsequent weeks the Luftwaffe showed the British the locations of all its other coast radars by surrounding them with masses of barbed wire, which showed up beautifully in reconnaissance photographs. (The Bruneval raid alerted the British to the exposed position of the vital TRE, and it was accordingly moved to Malvern.)
Of course, by this time Würzburg was an old set, fast being supplemented or replaced by the Gigant variety. For better early warning, Freya was being supplemented by a huge new radar called, appropriately, Mammut. This had an aerial like two bedsteads back-to-back measuring 45 feet high and 90 feet wide, with electronic switching through an arc of 100°. This rapid switching, which was much later to become a feature of night-fighter radars, allowed the beam to sweep across the sky while the aerial stayed fixed. The other early-warning set was Wassermann, with a rotating aerial about 130 feet high and 30 feet wide. Both the new sets had narrow beams enabling them to see aircraft 150 to 200 miles away. In most respects, they were superior to Britain’s prehistoric CH system, though neither was a patch on the monster MEW (Microwave Early Warning) radar developed in the USA. This was first installed at Start Point, Devon, where it could see every aircraft in southern England and northern France on D-Day.