The first night fighters to operate in 9 Group were a section of 29 Squadron Blenheim Mk IFs which had arrived at Tern Hill in August. The detachment was only on loan from 12 Group and it was detailed to provide one aircraft at immediate readiness and another two at fifteen minutes’ readiness. No. 29 Squadron crews operated along a number of patrol lines, one of the first of which was the Mersey Blue, that covered the approaches to the River Mersey.
Under 9 Group it was superseded by the Holt Patrol Line which was routed along a series of datum points and towns. The patrols took the Blenheims on a south-westerly – north-easterly track, across Middlewich and Holt, between Liverpool and Manchester. The patrols were operated to protect the various approaches that the enemy used to attack Liverpool and Manchester but they proved to be no deterrent.
The three-seat Bristol Blenheim first flew in 1935 and was a technological quantum leap among RAF aircraft at the time. With a top speed of around 428kph/266mph, the Blenheim bomber was considerably faster than the 290kph/180mph Hind biplane it replaced and it could outrun many contemporary fighters. The first Blenheim fighter, the IF, was proposed as a long-range fighter that could escort bombers over hostile territory and also carry out ground attack missions of its own. Around 200 Blenheims were modified for these fighter duties, additionally armed with a gun pack beneath the fuselage consisting of four machine-guns. The type had first entered service in December 1938 and by September 1939 there were 111 Blenheim fighters in use with the RAF. Unfortunately the Blenheim could not match the performance of aircraft such as the Messerschmitt Bf109 and so many became nightfighters, ultimately carrying the new and highly secret airborne radar.
Even before the IFs were equipped with radar they achieved some nighttime victories – in June 1940 No. 23 Squadron destroyed a Heinkel 111 bomber over Norfolk. The first ever radar interception came in late July when a Blenheim IF of Tangmere’s Fighter Interception Unit destroyed a Dornier Do 17 near Brighton.
The pioneers of the Blenheim nightfighters were a flight of No. 25 Squadron who were in fact the first unit in the world to operate radar-equipped nightfighters. But Blenheim fighters continued to operate in daylight too and as late as August 15, 1940, during the Battle of Britain, No. 219 Squadron were in action against a German raid on north-east England. Between November 1939 and March 1940, RAF Coastal Command also operated IFs providing top cover for shipping. The Mark IVF was again a long-range fighter version of the Mark IV bomber, carrying the same gun pack. Around 125 served with Coastal Command, providing shipping with air cover, as had the IF. In ApriI 1940 a pilot of No. 254 Squadron shot down a Heinkel 111 that posed a threat to British ships off the coast of Norway.
In October 1938 the Air Ministry belatedly recognized that there might be a need for a long-range escort fighter with two engines. Lord Dowding told me this was principally because of the existence of the Messerschmitt Bf 110, which required a response. It ordered an emergency ‘crash programme’ to convert Bristol Blenheim light bombers as the only suitable aircraft available. Even this, be it noted, had nothing at all to do with night fighting. In 1938 the Blenheim was still regarded as a modern machine of outstanding performance; it could, for example, easily overtake the Gladiator single-seat biplane fighter, which had entered service only the previous year and was fast becoming the RAF’s main day (and night) fighter. Contracts were placed with the Southern Railway’s Ashford (Kent) works for an eventual total of over 1,300 gun packs, each containing four of the reliable Browning machine-guns which had been designed in the United States as a 0.300 in calibre weapon in 1916, used in the First World War, adopted by the Air Ministry in 1934 (in the absence of any modern British gun) and put into production by the BSA company following a licence agreement of July 1935. The four-gun pack for the Blenheim was bolted on under the bomb bay, in which were stored the four belts of ammunition each containing 500 rounds, amply meeting the requirement for twenty seconds’ continuous firing (see drawing above). The railway workers met every tight schedule imposed upon them, and also delivered several other items making up each aircraft kit. Other companies supplied the reflector sights and extra armour to give some frontal protection. The first contract was for kits to convert 200 Mk I Blenheims into the Mk IF fighter, and these began to enter service in December 1938 with No. 25 Squadron. Later contracts covered the conversion of the long-nosed Mk IV into the IVF. These makeshift long-range fighters were among the busiest aircraft in the RAF during the first two years of war.
Had the conversion not been ordered it is difficult to see what aircraft might have been used to carry the first operational AI radar installations. By 1939 the AI Mk I, with its heterogenous assortment of hand-fitted parts and wiring, which made no attempt to comply with any airworthiness standard, had been replaced by AI Mk II, which was broadly similar but was a properly designed installation which had been ‘productionized’ and improved by industrial firms. Main contractor for the transmitter was Metropolitan-Vickers, while the receiver was made by Pye (the chassis was based on a commercial TV set). A main subcontractor was A.C. Cossor. The installation weighed about 600 lb, and exerted only a slight detrimental effect on the Blenheim’s performance. At high power settings an AI-fitted Blenheim could usually reach 250 mph without trouble, which was ample to overhaul a lumbering Heinkel. On the other hand, it was much too slow to catch a Ju 88 that had become jittery and ‘poured on the coals’; even a cruising Ju 88 or Do 17 was hard to overtake. One felt that lack of a properly designed long-range or night fighter was something that might have been avoided.
Inside the Blenheim Mk I there was a distinct absence of room, which must have been worse with radar fitted. The pilot sat on the left of the nose, his seat, instrument panel and controls filling exactly half the width of the fuselage. If a dual instructor station was added on the right, the two pilots rubbed elbows. Most of the nose structure comprised twenty flat Perspex panes, which reflected illuminated items in the cockpit and interfered with external night vision. The highly rated Mercury engines were reliable, and seldom put the pilot in the rather marginal position of having to go round again on one, with landing gear and flaps down. Behind the pilot, facing aft, the AI operator fiddled with his box of tricks and tried to get intelligible blips so that he could tell the pilot which way to steer (he had to remember which way he was looking and not muddle left and right). In an AI chase he had no time to clear stoppages with the guns or do anything but look into his viewing hood, which was retained even though it was dark in the middle of a Blenheim at night. Usually a third man, the observer, was also carried; he manned the turret, with a Vickers K gun, and tried to live up to his name.
The world’s first radar-equipped night fighters resulted from a secret minute from the Air Staff dated 17 July 1939, calling for the fitting of radar to twenty-one Blenheim IF long-range day fighters ‘as quickly as possible’. The document continued, ‘A requisition is enclosed incurring an expenditure of approximately £4,650 to cover AI transmitters, receivers and associated equipment.’ (Though the comparison could hardly be more meaningless, twenty-one AI radars in the year 2000 would be likely to cost something over £10 million, provided they were already in full production.) The AI II sets were delivered quickly by Pye and Metrovick, and much of the material needed for installation was bought by local purchase by the RAE at Farnborough, which did the installation with the help of engineers from Bawdsey. AMRE itself was being torn apart ready for emergency evacuation, in a badly planned way, to Dundee and other locations.
Despite this upheaval, AI staff were able to help at Farnborough, and deliveries of improved Blenheims began to 25 Squadron on 31 July 1939. When the Second World War began on 3 September a total of fifteen aircraft had been delivered, and the last of the twenty-one was in service by the end of September. All these aircraft had a much more sensible interior arrangement. The useless dorsal turret was removed and replaced by a hatch. In theory a Vickers K machine-gun could be fired from this for rear defence, but I cannot imagine this ever being done. The important thing is that removal of the turret enabled the bulky and heavy radar to be moved aft to preserve the aircraft’s centre of gravity in the correct place. Its viewing scopes could now face aft, so the operator no longer had to think in terms of mirror images before giving directions to the pilot. There was no longer a need for a third crew-member.
In November 1939 three of these Blenheims were delivered to a special flight of 600 (County of London) Squadron, Auxiliary Air Force, at Manston. Here they did special trials, in addition to crew training, and in early 1940 formed the nucleus of the Fighter Interception Unit (FIU), which grew rapidly and built up a collection of many kinds of fighter in its task of solving every sort of interception problem. Other early Blenheim night fighters were issued in ones and twos to existing Blenheim IF squadrons, beginning with 25, 29, 141, 601 and 604. The RAF grapevine buzzed with talk of ‘Magic Mirrors’ – talk which, as is traditionally the case with new RAF equipment, became slightly soured. The Magic Mirrors were difficult to use, and results in training flights varied from poor to non-existent. A basic snag was that the target was seen, if it was detected at all, in a position relative to the aerials on the fighter. If the fighter banked, the apparent target position moved in response, and it was eventually judged that ordinary Blenheim crews would have little success unless their aircraft was flying straight and level. Another of the many problems was that, unless the target was within a 20° cone ahead of the fighter, the aerials gave ambiguous indications; for example, the target could be to the upper left, or at the same angle to lower right. It was also impossible to get clear indication unless target range was between 5,500 and 1,200 feet, and the equipment was almost useless at heights below 8,000 feet because of the ground return. From the start of the war low-flying aircraft, mainly He 115 seaplanes, had laid mines by night in the Thames estuary and elsewhere round the coast. Early AI radar was completely useless in trying to intercept them, though some crews tried to spiral down from directly above. Bearing in mind that most crews were not very experienced, and still found it hard to add two and two correctly while trying to navigate on a pitch-black night, it is easy to see that successful radar interceptions even at high altitude proved consistently elusive.
By October 1939 some of the difficulties had been at least partly rectified with AI Mk III, which had first flown in August. This had similar circuitry but a new aerial system which gave fewer ambiguity problems. The transmitter sent its pulses from a pair of swept-back dipole aerials, looking like a harpoon, on the Blenheim’s nose. Two similarly inclined dipole aerials well aft on one wing picked up reflections from the target and sent them to the elevation circuitry to show whether the target was above or below. Two pairs of plain vertical dipoles on the outboard wing leading edges did the same to indicate azimuth, left or right. The different signal strengths received in the various aerials made the bright blips grow longer or shorter, and displaced above or below, or to left or right, of the time-base centrelines on the observer’s display scopes. It still took a long time to get repeatable and reliable operation, or to interpret the indications correctly.
On 14 February 1940 an Air Ministry appreciation tried to look on the bright side: ‘In the general disappointment over the behaviour of AI Mk II and III it is possible that the limited but very real advantages of this equipment have been overlooked.’ The implication is clear: there were influential people incapable of seeing beyond the existing situation who were calling for AI radar to be abandoned as useless. Had they won the day it would have been serious for Britain, and later for the United States.
After the original crash programme to equip twenty-one Blenheims with AI Mk II, all AI installation was done by 32 Maintenance Unit at RAF St Athan, Glamorgan (South Wales). St Athan was a large base, and the arrangement was at first ideal because, after a disastrous few weeks in Dundee, AMRE was again moved, to St Athan. Here more than sixty Blenheims were fitted with AI Mk III during the first six months of 1940. All stripping out, rewiring and preparation with brackets and aerials was done at St Athan, but many of the aircraft were actually fitted with the transmitter and receiver at FIU, which moved from Manston to Tangmere. FIU organized training courses for aircrew, and the handful of crews that could claim to be proficient began the practice of continually visiting the operational squadrons. Night fighting was a new technique, which made the most severe demands on crews. More than in any previous type of warfare, the problems, the techniques and the equipment never stayed the same but were constantly changing.
During that long, hot summer of 1940 hundreds of night missions were flown over southern England with AI radar, many of them in the face of the enemy. Success was conspicuously absent, and combat reports tended to revolve around all the usual snags: ‘low oil pressure . . . ASI u/s . . . had to switch off right engine . . .’, plus ‘hopeless intercom’ (now of vital importance to get near the enemy at all) and new ones peculiar to night fighters: ‘severe shock as I touched the firing button’, and ‘interception abandoned when the AI set started to burn’. In the first of five interceptions on the moonlit night of 18 June the pilot of a Blenheim who had seen an enemy bomber visually was shot dead by a short burst from the Heinkel’s dorsal gunner. Five He 111s were downed on that night, but all by day fighters; nearly all the Luftwaffe losses on those summer nights were caused by Spitfires, and a few Hurricanes, which went up hoping to catch someone held in a searchlight beam. Then at last, on 22/23 July, a Do 17 was shot down after a true AI-directed interception. The Blenheim was flown by F/O Ashfield, with P/O Morris as observer and Sergeant Leyland as AI operator. Ashfield’s combat report was tantalizingly brief, commenting in one sentence on how they were hit by debris from their victim and then discovered that they were at a low altitude in an inverted attitude.
Could a crew of three really get upside down without being aware of it? The answer is, emphatically yes; an AI chase took every atom of one’s conscious attention. Irrespective of whether the AI operator had clear unambiguous blips or a maddening flickering fuzz, trying to decipher the true position of the target and pass steering commands to the pilot was more than a full-time job. There was no chance of attending to anything else. It was the simplest thing in the world to make one’s gyro instruments topple, and in the cold, clinical concentration of closing for the kill I almost believe a wing could come off and not be noticed.
In the midst of all the excitement the few really great brains involved were always able to spare a moment to take a broad look at the problem. It was Tizard who made sure that one point of fundamental importance was not overlooked. In May 1940, in writing an appraisal of the use of AI radar, he commented, ‘We have insufficiently considered its use by day, especially in cloudy weather.’ Tizard never ceased to prod the Air Ministry into giving the most careful consideration of every new idea that was not openly ridiculous. By 1940 sound location was fortunately dead, but infra-red (heat) methods were very much alive, and an IR detector was flight-tested. The brightest IR team was led by young Dr R.V. Jones at the Clarendon Laboratory, Oxford. Director of the famed Clarendon was none other than Lindemann, whose antipathy for Tizard was equalled only by Tizard’s for him; as Tizard had for over twenty years been the only rival to Lindemann’s claim to be most senior defence scientist, the problem may be self-evident. Watson-Watt managed to sidestep political feuding, to his own and radar’s benefit. It was Tizard who suggested that Jones might leave the Clarendon, though in 1937 his IR detector had sensed other aircraft at a range of just over 1,500 feet, and was easier to package and use than AI radar. Maybe Tizard was gifted with foresight to see that IR would for many years to come be thrown off the scent by fires on the ground, by the Sun and even by sunlight reflected from lakes and rivers. It was almost certainly a correct decision in 1937 to drop IR detection, though the technique returned in the 1950s, as will later be related.
Other techniques included searchlights and aerial mines, as well as an increasingly long list of impractical suggestions helpfully sent in by the public. As noted earlier, use of as obvious an idea as the airborne searchlight simmered during the First World War (but was never actually used) and emerged again with the sudden mushrooming of air defence in the late 1930s. Most of the airborne-searchlight effort prior to the outbreak of the Second World War comprised paper studies and argument, whereas reason suggests that it would have been more sensible to do a few cheap experiments and see if the idea worked. Instead, little or nothing was done until the night blitz was actually hitting the country in the closing months of 1940, as will be recorded in the next chapter. The same is true of almost all the other ideas, including mines. It was even true with the fundamental fact of how far a night-fighter pilot might be expected to see at night. This highly variable factor was self-evidently one that demanded the most carefully designed scientific research in an attempt to get meaningful numerical results. Instead the Air Staff, Air Ministry, the scientific committees and even night-fighter pilots did nothing but argue – quite literally a case of heat overcoming light. The CH system’s limit of accuracy of between three and five miles was much too far to be of any use at night; Dowding said, ‘It might as well be fifty miles.’ Hence the urgent and crucial need for an additional sensor – Churchill always called it a ‘smeller’ – carried in the fighter. Despite Jones’ neat IR installation, it was clear that it would be AI radar or nothing.
In the closing months of 1939, while the AMRE research team was uprooted yet again and set up shop at Worth Matravers, near Swanage on the Dorset coast, their masters in the Air Ministry became increasingly concerned about the basic AI radar problem of minimum range. AI Mk III had a maximum range of two miles (say, 10,500 feet) and a minimum range of 1,000 feet, though in those days AI was temperamental and actual results were less predictable. Frankly, even AI Mk III was pretty useless except for the vital task of training crews. For the stern test that could come at any time there was an overwhelming need for an improved AI radar with minimum range as near as possible to 300 feet, and with clearer and more positive left/right up/down indication than the ‘squint-eyed’ Mk III. With the coming of war, the whole British defence scene had changed dramatically. Radar, previously the secret preserve of a tiny team of ‘back-room’ workers, suddenly gathered into its fold fresh manpower by the hundred and soon by the thousand. Watson-Watt personally scoured the universities to scoop up bright talent and open their eyes to the scarcely believable facts that in a government defence laboratory it was possible to find academic freedom, a most enjoyable atmosphere, and technical problems as gripping as any posed inside ivory towers. Industry, too, was harnessed in the biggest possible way to provide brainpower and manufacturing capacity.
It was mainly the newcomers that were to make the dramatic breakthroughs in AI radar. Everything – minimum range, inability to focus into a narrow beam, and inability to intercept the low-flier because of the ground echo – kept emphasizing the need for much shorter wavelengths. Nobody knew how to generate enough power at short wavelengths, but in fact one major hurdle had been crossed back in 1921 in Schenectady, when Dr Albert W. Hull, of the US General Electric Company, had described a novel valve he had devised and named the magnetron. Many workers improved it during the inter-war years, but in essence it remained a resonant-cavity device like an organ pipe or other wind instrument. Unlike almost every other oscillator the magnetron could generate energy at fantastically high frequencies, with wavelength down to a few centimetres; but power was still very small. Much later a quite different valve called a klystron was invented, mainly by W.W. Hansen at Stanford (who devised the crucial part, the rhumbatron) and the Varian brothers. In the klystron the main structure is a special CRT, whose steady pencil-beam of electrons is turned into a succession of intense bunches by one rhumbatron and then caught within another. Both the klystron and the magnetron could generate waves so short they were called microwaves. By the end of the 1930s brilliant workers at MIT and Bell Labs had developed the basic theory of waveguides – essentially just very accurate metal pipes – for carrying microwaves and, by physically adjusting their dimensions, for tuning the waves to exact wavelengths. It was a new and exciting field, pioneered in the United States but, like normal scientific research in peacetime, freely published.
Shortly after the start of the Second World War Watson-Watt gathered some of his best captures from the universities and – after assuring them that, though they had arrived on ‘the Shanghai express’ they had return tickets – asked them to think about microwave radar. Some of them knew enough about the subject to know that it could not be done, except with uselessly feeble power. A few others thought it worth chasing, but doubted that anyone could build a receiver. By the end of September 1939 two groups had used their return tickets, but only so that they could go back to work on the problem in their own laboratories. One group under Professor Mark Oliphant returned to the University of Birmingham to study transmitters. Another under J.H.E. Griffiths went back to the Clarendon Laboratory to work, in partnership with the Admiralty (under C.S., later Sir Charles, Wright, Director of Scientific Research), on the receiver. It was a mighty task. Every previous attempt to generate powerful microwaves had merely dissipated the energy into the atmosphere, or into heating the hardware (at least once it actually melted). No way was known of building any kind of practical radar on a centimetric wavelength. Indeed, almost a year later a VIP in the scientific world, leading a party of visitors to see the first demonstration ever given of what can fairly be called ‘modern radar’, summed up rather loudly by saying, ‘Centimetric radar is for the next war.’
In the first month of war such a comment could not have been disputed, because nobody then knew how well the shanghaied men from Birmingham would do. In the meanwhile conventional 1.5-metre AI had to go ahead with all speed. There were quite suddenly a succession of minor breakthroughs, the greatest of which was the development of a new modulator by A.D. Blumlein of the His Master’s Voice gramophone company (the electronics giant EMI). This dramatically cut the time duration of each pulse, and overcame the problem of overlay at the receiver by the direct pulse from the transmitter. The General Electric Co. (no relation to the US giant) began its radar career by producing a smaller yet far more powerful main transmitter valve, the Micropup, giving 10 kW on 1.5 m (190–195 MHz). It was hardly bigger than a household filament light bulb. W.B. Lewis, an AMRE newcomer, achieved a breakthrough with minimum range. The resulting radar grew to have little left of AI.III save the aerials; receiver gain and time-base deflection were increased, and the boxes were so arranged that a single technician could adjust settings with a screwdriver and simultaneously watch the tube-faces (previously he could not do both). The result was AI.IV, the first AI radar that could give real results in Service hands, and the only one in quantity service until late 1943. Provided that a target was within a 40° cone ahead of the fighter, its direction could be indicated unambiguously within 10°. Straight off, in September 1940, Pye was given a contract for 600 sets, with major assistance from EMI.
Thus, during the crucial summer of 1940, there were already in Britain three generations of AI radar, discounting AI.I and II. They were separated in timing by weeks rather than years, yet such was the pace of development they were utterly dissimilar. Mk III was primitive, the minimum that could reasonably be supplied to the RAF. Mk IV was better, yet still a 1.5 m set with all that wavelength’s inherent shortcomings. Still in the laboratory was the new generation using centimetric microwaves. (Incidentally, the FIU was very proud of the fact that, despite not being an operational unit of the RAF, its crews shot down the first enemy aircraft to be destroyed by every type of AI from Mk III to Mk X.) But despite all this work by the ‘boffins’, the night-fighter strength of the RAF was woefully small. Almost the whole establishment of Fighter Command comprised Hurricanes (about thirty squadrons in mid-August 1940, despite terrible losses in France) and Spitfires (about nineteen squadrons), few of whose pilots had even tried flying at night, and which could find night raiders only by chance. There were a few squadrons of Gladiators, some of which had done a little night flying, and an embryonic night force of Blenheims and Defiants. In September 1940, when the night blitz started in earnest, Fighter Command had, for all practical purposes, six squadrons of Blenheims and three of Defiants for night fighting. About one-third of the Blenheims had AI.III, and there were still a few AI.II installations.
This modest night-fighter force would still have been of little use without four further technical developments which formed vital links in the chain of aerial defence. One was a grand design called GCI (ground control of interception) which owed much to W.S. Butement, who had proposed a 50 cm naval radar at the Signals Experimental Establishment as early as 1931. GCI included CHL (chain home, low) and gap-filler radars to cover the lower airspace, but the main item was a radar giving a picture showing the positions of fighter and bomber. The second essential was IFF, already mentioned, the earliest example of so-called secondary radar. Fitted to all friendly aircraft, it was triggered by the defending radar pulses to send back an enhanced and specially coded reply. Thus, when ‘interrogated’ by either a ground station or a night fighter, the ‘friendly’ would automatically show up on the display screens in a characteristic way, without its aircrew even being aware of what was going on. A ‘hostile’, on the other hand, would not know the IFF code, which was constantly being changed, and thus its radar ‘blip’ would be suspicious. However, it could not just be shot down without visual identification, because it might be a ‘friendly’ with its IFF unserviceable, or even just switched off. Over the years IFF, like ECM, was to grow fantastically in complexity and cleverness, and to give rise to the modern field of secondary surveillance radars (SSR) and transponder beacons. The third new development was a related system of Racons (radar beacons) placed in a chessboard pattern over southern England to give an equally distinctive blip on night-fighter radars (AI.IV onwards) and thus help the fighter to return safely to base.
One could overlook the fourth link in the chain, unless one had been a pilot at the time. Aircraft had previously used HF radio, which suffered from ‘static’ and speech distortion so badly that to a layman any received message sounded like unintelligible gibberish. Even professional aircrew often had to request, ‘Say again’. In 1940 VHF (very high frequency) radio arrived with marvellously clear speech just in time to play its vital part in the night battle. The fact that it had shorter range was of no consequence. With it came a standardized GCI language. Today one is amused, but in 1940 the GCI command, ‘Flash your weapon’, caused not a trace of a smile: it came at a time of mounting tension in the chase, and meant ‘switch on your AI radar’. ‘Increase speed’ was partnered by ‘Throttle back’, and at the end of the interception the crew would be told ‘Darken your weapon’.
It needed all these new inventions and techniques to construct an effective scheme of night defence. It is unfair to describe it as sheer luck that it all came together in the autumn of 1940; it was planned with great care and forethought, and had it not been for the fact that the Air Ministry Works Department – responsible for civil engineering and buildings – stubbornly resisted every attempt to substitute speed in place of perfection, almost the entire scheme could have been operational by the beginning of the war. For the first time in history it enabled a country to wait until hostile aircraft were approaching, and then send fighters where they were most needed. The only shortcoming was that the GCI system alone could not put fighters in visual contact with the enemy at night. Modern air traffic controllers will see the problem only too well. Their job today is to keep aircraft apart; in 1940 the task was to bring aircraft together, and to do it with primitive radars giving very ill-defined blips that flickered and jumped and sometimes just disappeared for no obvious reason. Visual contact at night meant a few hundred feet, whereas the best accuracy a really good GCI controller could hope for in 1940 was more than three miles. Even then, the fighter had to be going in the same direction at the same height. There were countless other snags, not least of which was the fact that the 1940 controller himself had little experience. Until many months had been spent sifting unsuitable people, he often lacked an understanding of the fighter pilot’s problems, and often also lacked the right kind of confident patience and encouragement that was vital to proper teamwork. Controlling was an art.
Few indeed were the people in night fighting who in 1940 had any artistry, experience or even confidence. Come aboard the Blenheim night fighter of F/O (later Air Commodore) Roderick Chisholm, as he tried to learn basic steps in the new trade at Middle Wallop that summer:
I was kept waiting, signalling for permission to land but ignored, for about half an hour. I was anything but composed, and when a turn proved too much for the directional gyro, which spun, I also lost my sense of direction. At last I was given a ‘green’, but the dim pattern of aerodrome lights made little sense by this time, and my approach to land was not aligned with the flare-path, whose direction I understood too late. I had to go round again. The wheels and flaps on a Blenheim came up rather slowly, and by the time I was ready to start a circuit I knew that I was several miles from the aerodrome; but I could not picture my position, and the lights which I could see did nothing but confuse me. I was flustered, and the situation suddenly got out of hand. I did not know where I was; therefore I was lost. A feeling of panic came over me, and I could not think of anything except getting down somehow on to terra firma. It must be this paralysis that causes the inexplicable night-flying accidents. It took a great effort for common sense to overcome this one instinct, which seemed still to work, to return to earth as soon as possible; but slowly this happened and item by item things were checked. What height? What speed? Climbing or diving? Where was I likely to be? Each of these checks, usually done instinctively and instantaneously, now needed a special effort . . .
There’s a beacon – what’s it flashing? – dot something, missed it – climbing too steeply, must level out – now, where’s that beacon again? – get the beacon paper (it’s too flimsy) – where’s the torch? – mustn’t get flustered again – there’s plenty of time – climbing again, must level out – Andover VL, Wallop DA – now, where’s the beacon? – there it is, think clearly, read it slowly – looks like a V then L – read it again, there it goes again: it’s Andover – about 240° for Wallop – settle down, align the gyro – steady – lights ahead – a beacon – flashing DA, that’s Wallop – this is simple – I could go on, I’d like to go on now – this will be a joke tomorrow.
I doubt that there is a single pilot who does not have similar memories from his early days when it all seemed to be just too much. Flying by night could bring the feeling flooding back – often fatally – to pilots who, like Chisholm, had long experience in daylight. For almost a year such men tried to master the Blenheim by night, while an experienced few operated AI-equipped aircraft at FIU and, increasingly, in the squadrons. Not many German aircraft came over at night to begin with, but FIU mounted night patrols to see if they could intercept any, and from 5 June 1940 AI was used in real chases of real targets. Time after time the enemy got away, usually because the harassed AI operator could make no sense of the erratic blips of the shaky Mk III and was incapable of giving his pilot clear steering commands. The only exception was Ashfield. His was the first of a handful of victories gained by the NF Blenheims. Courageously flown for long, cold and exhausting patrols night after night, they simply lacked a good enough AI set to complete the chain of defence. Unlike most other fighters, including the Defiant, their guns could be fired on the darkest night with no flash problem. Performance, however, was marginal, even after the Blenheims had had their turrets removed in September 1940.