Back to main index

As tribute to those who lost their lives
On the 60th anniversary of the death of Alan Blumlein and ten others in the crash of Halifax V9977,
I have published here for the first time extracts from Chapter 9 and Chapter 10 of my book
in which the specific details of the crash and the subsequent investigation are outlined.
I hope that you find this of interest.

Chapter Nine (extracts)

H2S - The coming of
centimetric radar

The Origins Of Centimetric Radar

It must be remembered that the urgency with which centimetric radar had been developed was driven by the pace of the scientists working to develop what they saw as a logical improvement on the existing airborne interception radar system. Centimetric AI had been driven forward because of the poor results that the RAF night fighters were having against German night bombers during the Blitz. The knowledge of the poor Bomber Command results had not been commonly known at TRE, nor had the subsequent investigations into what was causing them or how they could be improved. Bomber Command, as mentioned earlier, had never approached TRE for any targeting or navigational bombing aids as other services had done. Because it was assumed that the bombing results over France and Germany were acceptable, even morale boosting to the British public, nothing had been done to speed up the development of centimetric radar in relation to bombing aids - though it could undoubtedly help. All that was about to change however.

Dee now wanted to organise a specific test to formulate the possibility of centimetric radar being used to scan the ground for echo returns. By modifying the rotation of the parabolic scanner to a fixed angle of depression to the horizontal, he felt certain that objects as large as a town would show up despite the ground scatter. Dee had installed, in August 1941, a double paraboloid helical scanner in another Bristol Blenheim Mk.4 aircraft (V6000), for flight trials on the AIH type of scanner. He was so certain that this method would work for the type of tests that would detect ground-based objects that, in the week following the Sunday meeting with Rowe, he asked Dr. Bernard J. O'Kane (who was a member of the GEC staff attached to TRE, and the project leader for V6000), and Geoffrey S. Hensby (the 23-year old radar operator), both of whom were actually carrying out the tests, to make the necessary modifications.

Hensby had been an engineering apprentice at the Royal Aircraft Establishment, Farnborough from 1934, to 1938, during which time he achieved a B.Sc. and M.Sc. as an external student at Birkbeck College. He was appointed 'Technical Assistant III' in July 1938, but chose instead to take up a DSIR studentship at Birkbeck in September, studying cosmic rays under H. J. J. Braddick. At the outbreak of war, Hensby had returned to RAE, and taken up the position of Junior Scientific Officer working on proximity fuses when they were in their infancy. Hensby joined TRE in November 1940, along with several other scientific officers from RAE, all needed to assist the urgent development of radar. It was at TRE in November 1940, that Hensby first met Sam Curran, who was immediately impressed with the young man, and asked him to come and work with him on solving the problem of driving the magnetron, which they did by introducing triggered sparkgaps. Hensby then turned his attentions to the problems of AI systems for night fighters, becoming instrumental in the installation of, and later flight-testing of aircraft fitted with experimental centimetric radar from Hurn airfield, near Christchurch. For a while, Hensby had also been aboard the test aircraft as observer during many of the AIS trials. As some of these flights were conducted over enemy-held territory, Hensby was given the protection of an RAF honorary commission and uniform.

Figure 9.01 & Figure 9.02- Bernard O'Kane (left) and Geoffrey Hensby (Courtesy of Sir Bernard Lovell)

On 27 October, O'Kane and Hensby visited GEC in Wembley, in order to obtain the agreement needed to interrupt the AIH tests, and to make the modifications to the scanner that Dee had thought necessary. This done, the TRE workshops then modified the scanner (which had been taken there on 28 October). The first flight test was carried out in V6000 on Saturday, 1 November 1941, and soon after take off, at about 5,000 feet near the Solent, radar operator Hensby reported to O'Kane that he could clearly see the large echo from the city of Southampton below. He noted especially that because of the absence of an echo return from the sea the coastline was shown in sharp relief. Encouraged by the success of the initial flight O'Kane and Hensby made subsequent flights at 7,000 feet over Salisbury Plain with photographic equipment this time to record the results. The photographs taken directly from the CRT clearly show the towns of Salisbury and Warminster, as well as the military encampments on the flat ground between them. When Dee was shown these photographs, allegedly before they were even dry, he apparently rushed them over to Rowe's office. With the wet photographs in front of him Rowe exclaimed, "This is the turning point of the war".

Naming The System 'H2S'

There are several versions of the story of how the new radar system came to get its name of 'H2S'. When the device had first been brought to the attention of R. V. Jones, by now Assistant Director of Intelligence and a close contact of Lord Cherwell's, he noticed that it had been given the code name 'TF' which, he felt inappropriate as it would be interpreted as 'Town Finding'. This was not specifically what the system did nor was intended to do, and Jones, decided to speak to Lindemann about it. According to Jones, Lindemann apparently remarked that it was, "...stinking that TRE hadn't come up with the idea for the device years earlier, and that an appropriate name for it would be H2S". H2S is the chemical symbol for Hydrogen Sulphide - a colourless, poisonous gas having the odour of rotten eggs!

There is another story of how the system got its name: Rowe said that Lindemann had chosen the name 'Home Sweet Home' because the system enabled one to locate and home in straight onto the target. That in turn became shortened to the name 'H2S' by TRE, though it is uncertain who actually abbreviated the new name first. It may have been Lindemann looking to find a simpler formula for the device, though more than likely it was the actual bomber crews first using the device, who shortened 'Home Sweet Home' to 'H2S'. In any case the name stuck. Rowe informed Churchill of the results of the flights made by O'Kane and Hensby on 20 November 1941. He was cautious enough to point out that a decision to use centimetric radar should only be made after serious consideration be given to the possibility of losing the equipment over enemy territory. Churchill was very enthusiastic about the potential for the new 'BN' system (Blind Navigation) as it had now been called (before H2S had stuck).

Over the coming two months, through November and December 1941, many more test flights were carried out in V6000 to determine which altitudes gave the best results. At this time Southampton, Bournemouth and Wolverhampton were all detected at ranges of up to 35 miles with the angle of approach immaterial, even the hangers of aerodromes could be detected with the radar. The tests were classified as 'Most Secret' by now, and the first murmurings about the consequences of losing a magnetron over enemy territory were voiced. This would be a subject that returned time and again as the possibility of losing the new oscillator and its secrets being discovered plagued those who had to decide whether or not to use it.

Figure 9.03 - H2S Scanner Beam (Courtesy of EMI)

Because of the potential disaster that could come from losing a magnetron, the tests were continued on the basis that the klystron valve might have to be used instead. They were subsequently carried out by O'Kane and Hensby using a magnetron to simulate the results that they expected to obtain from a klystron valve. Crucially however, instead of actually using a klystron during these tests, they actually reduced the overall sensitivity of the magnetron, which at the time showed "No sufficient reduction in range that would impair the usefulness of the apparatus". This observation would be destined to cause much trouble later on.

The next phase of the development of BN would be to evaluate its use at higher altitudes. The Blenheim that had been used up to now could not fly above 10,000 feet because the experimental equipment would flash over. A Beaufighter could fly at 20,000 feet for high altitude tests, but what was really needed were evaluation tests in the new series of bombers that were currently in development for Bomber Command. The four-engined Halifax, Stirling and Lancaster bombers were to replace the two-engined Wellington, Hampden and Whitley's as soon as they were available.

On 23 December 1941, the Secretary of State for Air, Sir Archibald Sinclair (who had succeeded Lord Swinton as Air Minister with the appointment of Churchill as Prime Minister on 10 May 1940), called a meeting with Lord Cherwell and directed that TRE proceed with the development of H2S for the new breed of bombers. Sinclair also instructed that H2S should be equally evaluated, in a simplified form, as a ground-mapping method for the identification of a target, rather than the range/azimuth indicator that the current TRE radar projects all used. Cherwell pointed out that such a system would be too complicated and take too long to develop. He suggested a compromise in which a split aerial system with a left/right display could be used to home in on a target. This he assured Sinclair could be developed much easier and would be adequate. Finally, the men decided that despite the promising results from the magnetron-based system, the klystron and the magnetron would need to be tested to evaluate which would be most suitable for the device. The Ministry of Aircraft Production had been instructed several months before that any project involving the klystron should be considered an additional project, and that nothing was to impair the production of the magnetron valve at any cost.

However, since there remained some doubts about whether the response from H2S were unambiguous related to specific targets, Cherwell insisted that "Further experimental flights were to be made immediately to determine whether signals obtained on the device could be definitely associated with specific ground objects." Of the other matters that were discussed, the issue of when the new device might be ready for active service was brought up. Cherwell and Sinclair both concluded that even with a joint development program involving TRE, EMI and GEC, H2S would be not be available before the autumn of 1942, but, once available, it was more than likely to be successful.

Immediately after this meeting, Dee, O'Kane and Hensby visited the Aircraft and Armament Experimental Establishment at Boscombe Down, to evaluate the three types of new aircraft to see which would be best suited for initial installment of the BN equipment. After looking over the Lancaster and Stirling, they decided that the Halifax offered more scope and alternative positions for the installation of the scanner than the other two aircraft. Therefore, it was decided that the Halifax bomber, then only just coming into service, would become the first aircraft to trial with H2S.

Rowe now needed somebody to head the BN development programme for TRE. He chose Bernard Lovell as director, because of his close association with the development of centimetric radar, not least of which was his work perfecting the parabolic scanner. Lovell, at first, was none to keen on the idea; In his book Echoes of War, he recalls how, at the time, he was busy working on a new feature of the improved AI system which would lock onto and follow an enemy aircraft (known as AIF): "On 29 December 1941, Rowe had summoned me to his office and told me I was to cease working on the lock-follow AI and take charge of the development of a new device, then known as BN (Blind Navigation) to help Bomber Command. I responded that I did not want to do this because I was anxious to get the AIF system into the Beaufighter. Dee was absent but when he returned of 31 December, he said I had to do this and insisted on driving me to the new aerodrome at Hurn, hoping thereby to lessen my resistance. Early the next morning I was again taken to Rowe's office whose patience was rapidly evaporating as he abruptly terminated my further objections by the terse statement that 'there was no alternative'. In that manner I was ordered rather than 'given charge of' a task that I did not want to do and knew little about"

One of Lovell's first tasks was to explain to Handley Page that the two brand new Halifax bombers that were to be delivered to TRE would need to have a Perspex cupola mounted on the underside of the aircraft fuselage, in order to house the parabolic scanner. Lovell had to face Mr. Handley Page, G. R. Volkert, his Chief Designer, and his design team on 4 January 1942, in the company of Bob King, also from TRE, who would be responsible for the fitting of the equipment into the aircraft, and A. W. Whitaker from Nash & Thompson who would fit the scanner. The three men had to explain that the modification to the Halifax needed to be made, regardless of the fact that Handley Page protested it would affect his aircraft's' performance which would be 'ruined,' and the bomb load severely reduced. Lovell remarked that it would surely be better to drop some of the bombs in the right place, than the entire load over the open countryside but, of course, he could not say anything about the nature of the work that they were doing at TRE, even to Handley Page at this stage. Other than give the specifications of the Perspex cupola, 8 feet long, 4 feet wide, 18 inches deep, there was little else Lovell could do to pacify Handley Page, who would just have to get on with the task.

It is remarkable today to think how quickly the aircraft were in fact delivered to TRE. Halifax V9977, the first of the two, arrived at RAF Hurn on 27 March 1942, just two and a half months after Lovell's visit to the Handley Page works, with the second aircraft R9490 arriving on 12 April, surely a testimony to the importance now placed on the task of perfecting H2S and the direct pressure being applied by Churchill on TRE.

EMI are brought in to the picture

Alan Blumlein and Eric White had first been informed of the project that would eventually become H2S during a visit to Worth Matravers in the spring or early summer of 1941. Though the vast majority of the staff and personnel at Worth Matravers had been moved to Leeson House at Langton Matravers (following the Heinkel incident in August 1940), there were still certain facilities kept at Worth Matravers, and it was here that Blumlein and White were to have a most extraordinary meeting. At this meeting was Frederic Williams and W. B. Lewis, It was they who informed White and Blumlein about the H2S project and then asked them if they would consider developing and engineering the circuits for the system. Naturally, Blumlein and White accepted the challenge on behalf of EMI and the team at the Research Laboratories. In the months that followed a vast amount of work was done developing these circuits, while A. F. Pearce at EMI also worked on the development and improvement of high-power klystrons. Following the meeting of Sinclair and Cherwell on 23 December 1941, it was decided that an immediate contract for 50 units of electronics for the H2S system would be placed on EMI. O'Kane was to bring the team of EMI scientists led by Alan Blumlein up to speed on the developments that had been made to that point with H2S; Lovell remarked at the time that he: "Did not, at first, anticipate any serious problems."

TRE were informed on 6 January, that a system capable of detecting a town at a range of 15 miles from a height of 15,000 feet would be acceptable, an immediate order to produce 1,000 units would therefore be placed upon the relative manufacturers. The device was not to be produced as a navigational system, but should be considered solely as a means of identifying and homing to a target. The klystron should be considered as a vital part of the system as the potential disaster associated with the loss of a magnetron-based system was considered too great at this stage. Magnetron development was to be continued however, with the highest priority in order to assess its potential further.

By 21 January, the Chief of the Air Staff, Marshall of the Royal Air Force, Sir Charles Portal, had informed Churchill that the progress on the magnetron system had been so satisfactory that a full development plan for H2S could now be formulated. Portal proposed that the existing contract on EMI for the electronics, on GEC for the valves and on Nash & Thompson in Surbiton and Metropolitan Vickers in Manchester for the scanners (both of whom had been originally sourced by Lovell when developing the scanner), should now be increased. The Ministry of Aircraft Production should draw up a programme with dates, for the provision of 1,500 H2S sets, to be manufactured by EMI, with a view to introducing the system as soon as possible. At the same time it was decided that Dr. Sam Curran should be the liaison officer in charge of all communications between TRE and EMI. The RAF would also have their own liaison team comprising an experienced captain and navigator; this team was headed by Wing Commander G. P. L. Saye with Group Captain W. E. Theak and Flight Lieutenant E. J. Dickie. They would act as liaison officers between the RAF and TRE/TFU.

As the evaluation of the two valves continued it became obvious that the tests should be split into two separate groups, one working on the klystron while the other concentrated on the magnetron. At a meeting at TRE on 26 January, a decision to conduct a series of experiments to determine which valve should be used for the H2S system was called for. The experiments, which, it was hoped, would take two to three weeks to complete, should determine finally which of the two should have the undivided attention of the manufacturing facilities. EMI, it was decided, should conduct the tests into the potential of using the klystron and, should it prove to be the more favourable of the two types, they would be contracted further to manufacture the electronic components. They had of course also been involved with other elements of radar development not the least of which was the GL project, also led by Alan Blumlein. The EMI team was assembled with Blumlein in charge, working with C. O. Browne and F. Blythen at Hayes, while F. R. Trott and M. G. Harker were to go to TRE and work on-site testing the equipment as it was manufactured and delivered. E. L. C. White would also spend some considerable time on site working with TRE. Despite the work that EMI had already carried out on GL and later AI, Blumlein's team had not really been working with klystron valves up to this time, except for a some early work carried out by John Collard on components for the 3 cm system. The Research Labs had not needed klystrons, as the frequencies used for AI Mk.IV & Mk.VI were obtainable with triode valves. Therefore, even though the cavity magnetron would seem to offer the best solution to the H2S problem, EMI were to be supplied with klystrons. It should be remembered however, that O'Kane and Hensby had initially reported that there should not be any reason to consider the use of a klystron with H2S as a reason to impair the device. It was tasked to the Works Designs department at EMI to manufacture the parts for the first H2S sets. These were then assembled and tested by the Research Laboratory (as, much later, the first 60 production H2S sets would be, when they came off the production line at EMI for the Pathfinder squadrons), in what had been the old auditorium where Blumlein had done some of the original binaural recordings and where some of the binaural films had been shot.

The major problem with the use of the klystron in the H2S device was that O'Kane and Hensby had made an error in judgement when concluding that the klystron would work as suitably as the magnetron valve. This was clearly not the case. Their deduction had come from reducing the power on a magnetron during a test to 'simulate how a klystron might respond'. Magnetrons however, working at much higher power outputs, could detect targets at 35-50 miles. With the klystron valve installed, initial results would yield results that only obtained targets at ranges of 5 miles or less. A single magnetron valve was given to EMI in order to assess the variances in their work with the klystron, though the majority of the evaluation tests on the magnetron would be carried out by O'Kane, Hensby and Lovell's team based at TRE. At the same time, work was also to continue towards the introduction of Gee which, by February 1942, had been developed to a stage where it was almost ready to be installed in aircraft in active service.

Pre Production Tests

When Churchill had given his express orders for the two Halifax bombers to be made immediately available to TRE/TFU, they were literally coaxed away from Bomber Command at a time when the air war was going very badly for the Allies. Aircraft losses in first few months of 1942, were increasingly high, and it was felt by many at The Air Ministry that few if any could be spared for research that might potentially risk the aircraft and crew, with no guarantee that any development might benefit the RAF. In March 1942, during the many raids that were carried out that month, nearly 200 aircraft were lost to enemy action. That figure, apart from the enormous loss of men and materials that it represented to Bomber Command, was also very near the maximum for aircraft production and crew training programmes to keep pace. Despite the very definite opposition voiced by some to the huge effort, not to mention considerable cost that was now flowing into the radar development programme, Churchill continued to champion it's cause. That was why the two Halifax aircraft were given to TRE with such a priority attached to them.

At EMI Blumlein was busy assembling the team that would work on the klystron development of H2S and bringing everybody up to speed on the new system and how it worked. The scientific team that he chose would consist of himself, Frank Blythen, Felix Trott and Cecil Browne. The engineering team, who would actually be responsible for the building of the first H2S units, comprised H. L. Oura and W. H. Cox. The work itself was carried out on the first floor of the Works Designs Department, Building No.1 at Hayes, in Middlesex. Curran liased directly with Blumlein who in turn reported to Lovell all aspects of the development progress.

At GEC the klystron and magnetron valves were already under contract to be manufactured although progress was proving painfully slow. Nash & Thompson were manufacturing the hydraulic scanner apparatus and Metropolitan Vickers were building an electrically driven version of the same. The assembly of the various elements of the system would obviously take some time, during which flights trials on the original magnetron prototype would have to continue. While the two Halifax bombers were being readied, another Blenheim Mk.4 (V6385) at RAF Hurn was used.

Figure 9.09 - The Metropolitan Vickers electrically driven parabolic aerial underneath Halifax R9490 in 1942 (Courtesy of EMI)

During one of the 20 or so flights that were carried out, the radar operator recorded that strong echoes were received from the aerodrome hangers as RAF Yelverton from a range of 30 miles. Although the visibility was so poor that the pilot didn't see the hanger with his eyes until only 2 miles away, with the heading he was given he could fly straight toward it.

At the English Electric works in Preston (where the first aircraft was being assembled) and at Handley Page (where the second aircraft was being assembled), the Halifax aircraft assigned to TRE were being fitted with their Perspex pods which were slung under the belly of the aircraft. This 'cupola', as Lovell preferred to call it, protected the delicate rotating scanner, but enabled the high frequency energy to penetrate and scan the ground without affecting the transmitted or received signals. During the time the development work on H2S had been progressing, centimetric ASV had also reached a stage where a provisional manufacturing contract had been placed with Ferranti in March 1942. It seemed obvious that many of the component parts of centimetric ASV and H2S would be the same and that, in order to save duplication, there should be some form of agreement where one installation could cover the needs of both systems.

This posed a problem because centimetric ASV had been designed entirely around the use of the magnetron, whereas H2S at this point in time still had development basis on the klystron, which was not considered effective enough for ASV use. The continuing debate over which of the two valves to use would now have to be resolved in order for centimetric ASV and H2S to proceed jointly manufactured. The problem facing the Air Ministry and the Government was simple. Which of the two valves performed the best? If it was the magnetron, could it be risked over enemy territory? The answer to the first question was becoming clearer by the day. Even though neither Halifax had been fitted yet with the development equipment, the flights being made by O'Kane and Hensby in V6385 were now definitely showing that their first observations regarding the results of the klystron were in fact in error. Not only did the klystron not prove anywhere near as effective as the magnetron results, the ranges they were obtaining and the altitudes at which the device detected towns was well outside the Air Ministry's minimum of 15 miles at 15,000 feet. It was becoming obvious that the magnetron was winning out over the klystron. Because this new system would effectively make any aircraft fitted with it independent of any ground control station for target identification, it would be limited only by the range of the aircraft itself. Exactly what Bomber Command needed. This then, left the dilemma of whether such a secret device should be risked in an operational aircraft over enemy territory. What if an aircraft was captured and the equipment fell into enemy hands?

The last thing the security conscious authorities wanted was to give away any advantage that might have been gained to the Germans. For this reason they favoured the klystron system as it was considered technology that the Germans already knew of having been published in detail as far back as 1939. True, it was not as good as the magnetron system, and it was significantly lower in power output although it worked in roughly the same frequency range, but could it be risked? Could a captured magnetron system be dismantled and copied by the Germans? If this were the case, then the German night fighter defence system would be able to match Bomber Command during their night raids. The potential losses in aircraft and crew would be enormous.

It was decided to see if the device could be destroyed with the use of a detonation charge that the crew could activate in the event of being shot down or imminent capture. However, even this had its problems. The main magnetron core was a solid copper block with eight bore holes around a central cavity through which the current was passed. It was nearly indestructible. Explosives experts at Farnborough devoted some considerable time and effort to the task of fitting a suitable destructor to the magnetron block and its connector systems. Even with the amount of explosive charge that it was considered safe to carry onboard an aircraft, a magnetron was extremely difficult to camouflage. During the destruction trials, a 2-ounce explosive charge was used (which blew a 10 foot hole out of the side of a captured German Junkers JU88 aircraft used for the test), the magnetron core was still identifiable, and worse, the fragments left over would have been, it was deduced, able to be studied to find out how it worked.

Clearly unsatisfactory the Air Ministry tried a method of explosive ejection and destruction in mid-air. But this also proved too dangerous for the aircrew whose responsibility it was to confirm complete destruction near the aircraft, but risked endangering the aeroplane from the recoil of the explosion. The idea of resorting back to the klystron oscillator to power the H2S system was reconsidered, but in the end it was events in Bomber Command that would really decide the outcome for the authorities.

By mid-1942, with losses over Germany getting close to the limit of 200 aircraft a month that had been set by Churchill, Dowding and Harris, the use of H2S, and the risk associated with its loss over enemy territory was considered justified - if it would considerably reduce the numbers of aircraft lost.

On 15 March 1942, the Ministry of Aircraft Production agreed that a proposal for one company to develop jointly the electronic components for centimetric ASV and H2S should be EMI, and that this development should concentrate on sourcing as many components common to both as possible. At this point they held short of finalising which of the two valves should be used as it was decided that this should be resolved with a short series of test flights in the new Halifax bombers which were now almost ready. TRE and EMI drew up a component list of interchangeable parts for centimetric ASV and H2S. The cupola, magslip, scanner, and cabling could be made common by changing the aerials of the H2S system which was already being made by EMI anyway; the power supply unit, receiver and timing circuits, control box and indicator unit could also be easily modified to work in both roles. In fact the only major differences at that time were three for centimetric ASV: the magnetron, a modulator and another power supply. For H2S, the differences were a klystron, and a modulator.

The Gramophone Company division of EMI were now given the task of putting this joint development programme into action which was known as H2S/ASV. There were to be 15 pre-production models completed by EMI as soon as possible, and the original manufacturing order placed on the company at the end of January for 1,500 H2S units, was now changed to 1,500 H2S/ASV units. The Air Ministry, TRE and EMI now decided that final decision as to which valve to use with H2S should be resolved by a short series of demonstration flights. This would determine the merits of both systems and then the authorities could make their minds up based on hard facts. By now the two Halifax's had had the modifications that Lovell, King and Whitaker had asked for and were ready to be delivered to RAF Hurn.


On Friday, 27 March 1942, Halifax bomber V9977, a Mk.2 built by the English Electric Company in Preston, was delivered to RAF Hurn. It was followed on Sunday, 12 April 1942, by Halifax R9490, also a Mk.2 but built by Handley Page. During the following 10 weeks, these two aircraft were to prove that Churchill's dogged persistence with radar development was worth all the effort expended.

Halifax V9977 was designated as the number one demonstration aircraft and was duly fitted with the H2S radar system which incorporated the magnetron oscillator, and being tested by the TRE personnel. The radar equipment was linked to a Nash and Thompson hydraulic scanner, essentially a parabolic aerial, suspended in the pod under the aircraft. The ground reflections were to be displayed on a circular display mounted inside the aircraft, positioned at the navigators station towards the front of the aircraft and immediately below the cockpit. From here the radar operator, peering into the PPI display could distinguish from the echoes received between flat countryside and taller built up areas with varying degrees of eminent light. The operator would view a sweeping clockwise rotary image with the echoes appearing as areas of light and dark sketches tracing out the countryside just like a map.

Halifax R9490 would also be fitted with prototype H2S radar equipment, but this time using the klystron oscillator system, being tested by the EMI team, and which was linked to a Metropolitan Vickers electrically driven parabolic aerial. The two aircraft were now ready to complete their flight tests and trials that would conclude which of the two systems worked the best and which eventually be fitted to the RAF night bombers of Bomber Command.

From its arrival on 27 March, until 15 April 1942, V9977 was in the hanger at Hurn having the magnetron H2S system installed. The first trial flight was to have taken place on 16 April, but there was an electrical problem with power to the H2S system supplied by an 80-volt alternator driven by one of the engines. The first test flight actually took place on 17 April, by which time R9490 had arrived at Hurn also, and was in the hanger being fitted out. During the flight, which was made by Lovell, Hensby, and O'Kane with Pilot Officer C. E. Vincent (who was a member of the TRE group responsible to Lovell), at the controls, an altitude of 8,000 feet was achieved at which point several good ground returns were seen on the CRT but, disappointingly, they were only able to detect these at a range of four or five miles.

Other flights took place in V9977, similar to that carried out on 17 April, but most of these also produced poor results compared to those that had been achieved in the Blenheim. This was mainly because they were carried out at higher altitudes which led to the conclusion by Lovell and Hensby that the scanning apparatus needed to redesigned and repositioned.

As if the scientists at EMI and TRE didn't already have enough to worry about with the poor results they were obtaining, the numbers of bombers that were now rolling off the production lines and reaching Bomber Command, meant that the possibility of a serious bombing offensive in the autumn now looked like becoming a reality. That offensive was reliant upon H2S being ready in time. As things stood at the end of April 1942, EMI didn't feel they could deliver more than 15 completed H2S units by Christmas - assuming they could get it to work properly.

Blumlein and the EMI team had been working non-stop during this period with Oura and Cox at Hayes, constructing the klystron version of the system. The work had taken longer than they had anticipated because of supply problems with certain components, having finished the work however they managed to assemble the apparatus in a coach and get it working ready for the drive down from London to Leeson House and RAF Hurn. They arrived at Worth on 14 May, and soon transferred the klystron equipment from the coach into R9490. Since arriving at Hurn on 12 April, this aircraft, like V9977 before it, had been undergoing ground flight check preparations prior to the installation of the H2S system. By mid May it was practically ready to begin flight checks, but these were cut short, as were the tests carried out in V9977, by the orders to move from Leeson House and RAF Hurn, to Malvern and RAF Defford.

Time To Decide - Then Disaster

The disruption that the move to Malvern must have caused is hard to underestimate. The move took place from Sunday, 17 May 1942, to the following Sunday, 24 May. Everything had to be stopped, packed in crates and shipped by road across England to Worcestershire. The two Halifax's had their H2S equipment removed by a combination of the EMI and TRE/TFU teams as well as the engineer fitters attached to TFU at RAF Hurn. V9977 and R9490 as well as the other aircraft attached to TFU were flown to RAF Defford where, upon arrival, they were subject to routine maintenance checks.

There had evidently been some problems with one of the engines on Halifax V9977 as it was decided at this time to replace it with another - on 25 May, the port inner engine (No.2) which had originally been a Rolls Royce Merlin Mk.20 was replaced with a Packard Mk.38. Though this was a time consuming procedure, it was not uncommon during the war, and the aircraft engineers at RAF Defford would have been able to carry out the operation in a matter of a couple of days. Between 25 May, and 1 June, all four engines of V9977 underwent a standard 30 hour inspection, and it was decided to adjust the tappets. It was also during this same period that the H2S units, both the magnetron version for V9977 and the klystron for R9490 were re-installed in the aircraft.

To add to the problems of the move itself, the rain and cold had been almost incessant during the entire time that everybody and everything had been on the road. There were major problems with accommodating everyone in Malvern and there was no provision for food other than to queue for ages in the rain outside the Winter Gardens Pavilion in Great Malvern while the Women's Volunteer Service cooked for the more than 1,000 staff attached to TRE/TFU. There were builders everywhere trying to erect accommodation blocks, laboratories, workshops - it must have been mayhem. Moreover, amongst all this, there was the constant pressure to get the H2S system working properly. The first of the re-scheduled test flights was to be carried out in R9490 on 2 June, Lovell and Hensby were on board with Vincent piloting the aircraft and Wing Commander Saye along to assess the klystron H2S unit's results. Unfortunately the flight was not very successful as the klystron once again proved that it was not powerful enough resulting in poor range, this time just four miles. Saye later complained after that flight, in a letter to the Air Ministry, that all he could make out was a 'snowstorm' on the CRT, that he could not determine this as a town or anything else for that matter. The klystron system was not working out. It was time to try the magnetron again.

The magnetron version of the H2S system had not been immediately installed back into V9977, instead it had been taken to the work benches that TRE were using on the top floor of the Preston Science School at Malvern College. Hensby and O'Kane had been working on the unit for some time since the flight of 17 April, which had been so disappointing. Now they thought they knew why. Blumlein requested that another series of test flights in V9977 with the magnetron H2S system should take place on the weekend of 6-7 June, during which time his EMI team could assess the progress made by Hensby and O'Kane. Blumlein, Blythen, Browne and Trott checked into a hotel in Tewkesbury a few miles from the aerodrome at Defford and spent most of Saturday, 6 June in discussion with Lovell, Hensby, O'Kane about the nature of the tests and how they were to be conducted.

On the evening of 6 June, Hensby and Lovell went up in V9977 as a pre-cursor to the following days flight which had been designated for the EMI team to assess the magnetron system. Lovell recorded that evening that the latest modifications to the radar produced good results. Gloucester, Cheltenham and several other towns were all clearly displayed at greater ranges than Lovell had ever seen before. All looked well for the tests the next day. By now, the weather had turned fine again. Sunday, 7 June, was a clear, sunny day when Halifax V9977 took off from Defford to conduct the tests with the EMI team of Blumlein, Blythen and Browne on board. Hensby and Vincent were also aboard as were the six flight crew.

It can only be assumed that the results of the tests would have been favourable, as they had been the night before, but we will never know. During the flight, Halifax V9977 developed a fire in the No.4 starboard outer engine, and in the course of trying to find a suitable location for an emergency landing, the aircraft crashed killing everyone on board.


Chapter Ten (extracts)

The loss of Halifax V9977

Secrecy & Speculation

The circumstances surrounding the tragic loss of Halifax V9977 and the 11 crew onboard have never been fully documented in an open publication. Needless to say, in time of war (and bearing in mind the sensitive nature of the work being carried out during the flight), much of the detail of the crash was hidden for years under a blanket of total secrecy. Some forty years after the crash, an investigation was carried out by engineers at the Radar Research Establishment at Malvern, but the document was intended as an internal publication and had been produced ostensibly to correct details which had originated from the crash investigation immediately after the flight, and which had subsequently been taken to be factual. The actual course of events and the reasons that led to the loss of the aircraft were therefore brought to the attention of only a few Ministry of Defence personnel, the civilian Air Historical Branch (RAF) archivists, and a select few who had personal or professional ties to the work that had been carried out by the team on that fateful flight. This inadvertently led to much speculation as to what actually occurred on that Sunday in June 1942, and how it was, that on a perfectly clear day, a nearly new aircraft with a total flying time since delivery of just 64 hours and 45 minutes, could fall from the sky killing all aboard and, in the process, rob the world of untold scientific genius.

Figure 10.03 - Halifax V9977 (Courtesy of Sir Bernard Lovell)

The Last Weekend - Friday, 5 June 1942 To Sunday, 7 June 1942

Eric White, Felix Trott and Maurice Harker all recalled the events of that weekend vividly when asked about it over fifty years later. On the Friday morning, Blumlein had called White to tell him that he, Browne and Blythen would be arriving that afternoon from Hayes to prepare for the weekend flight tests of the magnetron system in V9977, in order to compare the results that Lovell and Hensby had been seeing with the rather poor results the EMI team had been getting from the klystron valve system. White was instructed to inform the RAF personnel of the EMI team's arrival, and arrangements were made for authority to clear the flight itself. Preparations were made for Blumlein, Browne and Blythen to stay at the Swan Hotel in Tewkesbury, which was where White, Trott and Harker were all staying.

It is not known which of the three men actually drove the car up from London on the Friday morning, though it is quite probable that Blumlein did in either his own Ford V8 or an EMI staff car for which petrol rations would have been allocated. Following their arrival, and after a short introduction to the work in progress, Blumlein was asked by Maurice Harker if it would be alright if he might return to Pinner, near Watford, the following morning as his parents house had just suffered a burglary and he wished to try and help them sort it out. Blumlein told Harker that he thought it would be alright for him to return as there were plenty of extra EMI personnel at Defford with Browne, Blythen and himself there; and besides, Harker was finishing work on the AI Mk.VI at this time, and had not yet been brought into the H2S project. Blumlein told him to take the weekend, and to return on the Monday afternoon.

That Friday evening everybody ate dinner at the Swan Hotel and then all went for a walk as Maurice Harker recalled: "I was walking with C. O. Browne and F. R. Trott, with Blumlein and White a way off ahead of us. Well, after a while of this, it was a lovely summer's evening you see and not a bit dark yet, C. O. Browne says 'I've had enough of this', and he'd been walking with some stick or something, so, he proceeded to tap this stick on the ground to get White and Blumlein's attention or I think they would have walked another five miles, they were that engrossed in conversation. Anyhow, we returned to the Swan, and turned in for the night, and in the morning I left to return to Pinner and my parents house, while they were setting off to the airfield at Defford, and of course that's the last time I saw Blumlein"

The Saturday was spent in preparation for the flight trial the next day, with everyone returning quite late in the to the hotel in Tewkesbury. Felix Trott recalled that despite the day's work, there were still several pre-flight preparatory checks to be completed the following day before the aircraft could take off, but they did not expect these to take long at that time. It is also worth mentioning that in the Hotel that evening, Trott recalled the presence of a quite high ranking naval officer in the bar, having "…at least three stripes on his arm". Whether this man had anything to do with Defford or was there purely by coincidence, he did not know, but the man did stand out in his uniform that evening surrounded by civilians.

During Friday, after he had arrived at the airfield and been shown around V9977, Blumlein began to make some sketchy notes (which were probably meant as mere reminders to himself for when he had returned to Hayes), and continued to do so for the rest of the weekend. These hand-written notes concerning the tests to be carried out still survive in the vast EMI archives, and are, perhaps, the most poignant by him. Written in pencil on four pages which have been torn out of standard memo pad, and dated 5 June, 6 June and 7 June 1942, they give us a quite unique insight into the thoughts of Alan Blumlein as the EMI team prepared for the tests that lay ahead that weekend.

Figure 10.07 - Detail from Blumlein's hand-written notes on H2S made at Defford dated 6 June 1942,
the day before the crash
(Courtesy of EMI)

The Day Of The Flight

Sunday, 7 June 1942, was a bright clear summer's day, with patchy wisps of cloud 3/10ths and excellent visibility up to 18 miles with cloud at 3,000 feet. There was a light surface wind of 10 mph from approximately 300 degrees, the North, North West. It was an ideal day for flying, the kind of weather pilot's enjoy the most. The test flight to evaluate and compare the magnetron H2S system was scheduled for the morning, but because of several preparatory delays they didn't actually take off from Defford until 14.50 p.m. that afternoon. The object of the flight was to produce photographic images taken directly from the CRT of the ground echoes from the radar as the aircraft flew over the Severn Estuary, the coastline and Cardiff and Swansea.

The flight plan was chosen to highlight the prominent features of the area: large flat expanses on the Severn estuary, built up areas around Cardiff, Swansea and Newport, coastal plains and the sea, as well as several off-shore islands. This was part of the reason that RAF Defford had been so useful in the first place as a base for TFE/TFU as it offered close proximity to terrain of this sort. However, airspace was somewhat restricted by the 35 active airfield in the vicinity, most of which were actively engaged in the training programs for aircrew. This was why a Sunday was such an ideal opportunity for this kind of flight. There would be little chance of running into the various training missions taking place at the weekend, and the high number of airfields around offered the possibility of many emergency landing sites should anything go wrong.

Bernard Lovell had intended to fly on V9977, as had Sam Curran, the head of the EMI/TRE liaison team. Curran recalled many years later the events of that afternoon, "Clearly Geoff Hensby was required to go because of his in-built experience, and Brown and Blythen wanted to go, partly to see for themselves the conditions of operation as well as performance. With these three and myself, i.e. four civilians, plus seven RAF personnel, the total allowed, 11 in all, was reached. Then, out of the blue, Blumlein said to me he would 'love to go along' and I felt I had to volunteer to stand aside in his favour. Further, I said I would give him my flying suit as time had virtually run out, and I added that I had seen a system not unlike the one to be tried. So off went all 11 of them"

As Blumlein climbed aboard the aircraft he turned to Lovell and said, "Don't worry, we'll make it short eh?" Bernard Lovell had also given up his place on the aircraft to another of the civilian staff from EMI, either Cecil Browne or Frank Blythen (it is not certain which), with equally tragic consequences. Yet another member of the EMI team, Felix Trott, should have been on the flight, but being fed up with all the delays that had persisted before take-off he decided to go and get himself a cup of tea. This would have required walking to the far side of the airfield, so Trott borrowed a bicycle and headed off in search of his refreshment, evidently under the impression that the delays might continue for some time yet. When he got back from his refreshment, V9977 had left without him - surely the luckiest tea break in history.

Some sixty years later Felix Trott would recall the day vividly: "It was a lovely hot, fine, day; and I don't know why there was a vast amount of delay .... and things dithered and dithered and dithered, and it wasn't known who was going to fly and so forth. A couple more people from TRE turned up and they wanted to go, and there wasn't really any particular point in either myself or the other people who were stationed down there going on this trip because we could do it anytime. So, off they went. I got on the bicycle which we had down there and cycled round the perimeter track to the Officer's Mess and had myself a nice, cooling cup of boiling hot tea. "And when I came back they had gone, and one didn't expect to hear anything because there wasn't normally any radio communication anyway for a good hour at least. And I suppose, one began to get a bit anxious after two and half, three hours. And, how long in fact the station had known, I don't know; but eventually someone came down and told us that there had been a crash, which turned out to be miles away in the Wye Valley. And whether they had done the standard tests and said 'yes, this looks promising, now let's look at some more difficult terrain - I don't know. And I doubt whether anybody knows. The unfortunate thing about this aircrash was that the aeroplane that crashed wasn't the one that had the EMI experimental equipment on board, but was one that had the TRE equipment aboard, and Blumlein, Browne and Blythen, who were the three EMI engineers who were killed, had gone up by invitation to see how this equipment was working. But, it crashed, and that was the end of them"

From Take-Off To Cruising Altitude

Halifax V9977 started its engines and the crew went through their usual pre-flight safety checks. The ground crew pulled away the wooden chocks that held the wheels against sudden engine surges, and the aircraft slowly at first, building speed later, taxied onto the perimeter runways at Defford heading for the end of the main runway 22 (220 degrees heading) ready for take-off. During this time the scientific team would be busy checking that their equipment was working satisfactorily, probably for the umpteenth time.

At the end of the runway the aircraft halted and Berrington would have normally prepared to apply full power to the four Merlin engines to achieve take-off speed, but Berrington was not happy, there had been a power failure in the radar system which was driven from a generator housed in the S/O No. 4 engine.

Pilot Officer Berrington radioed the tower and explained that there was a problem with the power supply to the radar equipment, and could somebody come out to the aircraft and check it to see if it could be fixed or if they could find the problem. Obviously with such perfect flying conditions, coupled with the fact that the scientific team had already had several delays during the morning, Berrington wanted to avoid aborting the flight if at all possible. A call was made from the tower to Derek H. Moseley, the workshop manager of TRE Defford, explaining the problem. Moseley and Ron Hayman (whose rough notes have been mentioned earlier) were requested to go out to V9977, which was still standing at the end of the runway ready for take-off with all four engines running.

Hayman and Moseley went out to the aircraft and after an inspection, they very shortly discovered that the problem with the power generator for the radar system was being caused by a faulty cable socket connection on the front of the Voltage Control Panel (VCP), which was the electrical output to the radar. It would seem that the socket had not been properly screwed fully home on the front panel and this being quickly corrected, the radar functioned once again. Berrington was informed that all was well again and first Haywood and then Moseley climbed out of V9977 to allow it to continue on its way, thus Moseley became the last person in Halifax V9977 before the crash other than those killed.

With the power problem to the radar successfully overcome, Berrington obtained clearance for take-off from the tower and applied full power to the four Merlin engines, which thundered the aircraft down the runway. At around 14.50 p.m. Halifax V9977 cleared the runway at RAF Defford and maintained its heading ahead in the climb with the engines throttled back to 2,850 rpm at +9 pounds boost pressure, the standard climb setting of the four Merlin engines. At this steady climb setting giving 150 mph ascending at about 2,000 feet per minute the aircraft would have reached 10,000 feet to the West of Gloucester and thence by visual reference commenced its run over to the South West coast of Wales. Several previous routes for demonstration flights had taken the aircraft over an area between Gloucester and Kidderminster, but possibly because of the nature of the test and the need to compare the results to the klystron H2S system, the route was not to be taken on this day.

By the time V9977 took off on its last flight most of the members of the EMI team were convinced that the magnetron H2S system was the one that should go into production, and were obviously keen to impress upon Bomber Command the results of the equipment. For this reason the 'target' for the day would need to be distinctive and so a series of passes would be made over Cardiff and Newport to get good ground echoes of built-up areas, the Severn flats to demonstrate identification of coastal shoreline and then eventually back to Defford to evaluate the results.

Very little is known for certain of the exact route chosen for the flight as radio silence would have been part of normal wartime procedure. The navigators log and flight plans (assuming any were kept as Berrington may well have been flying 'visual') were all lost in the fire after the crash. However, V9977 was seen and recalled later by several witnesses on the ground (most of whom it must be stressed, were not called to appear before the Board of Enquiry as their accounts were considered to be non-verifiable). It is from these eyewitness accounts (accurate and inaccurate observations of aircraft on the day, many of which were probably not V9977), as well as deducing from the flying time and fuel remaining, that the following reconstruction of the events leading up to the moment of impact has been pieced together.

Having left Defford the Halifax would have required about 15 minutes to climb and reach its cruising height. However it would appear that V9977 flew at various altitudes on this day with ground reports (non-verified) placing it at between 1,500 feet and 15,000 feet for parts of the flight. The matter of the actual altitude flown has never been definitively established It is unlikely that it ever will be however, it should be pointed out that at altitudes above 10,000 feet the use of oxygen would be required. Berrington would probably not therefore have taken the aircraft to anything like 15,000 feet, and besides the purpose of the flight was to test equipment which was designed to look at the ground. An altitude of 15,000 feet would have been in excess of it's design limits.


It must be pointed out that the following altitudes reported on or about 16.00 p.m. that afternoon were from unidentified members of the public who presumably attributed their sightings to those of the crashed aircraft once they had heard of the accident. These matters were not followed by the Board of Enquiry as the sighting were non-verified and which aviation specialists of relevant experience have clearly identified as non-contributory to the events which actually took place. There were in fact several aircraft movements in the area at the time of the crash and any reference to low flying aircraft which were not trailing smoke/fire from the right outer engine were ignored. That aside, several so-called 'sightings' of V9977 have been used to position the aircraft in the moments up to and immediately following the engine fire commencement.

At about 16.00 p.m. the aircraft was seen from the ground circling a point some four miles North West of Usk in Monmouthshire at a height of approximately 1,500 feet. Some minutes later the starboard outer propeller was seen to stop, this would have been engine No.4. At around 16.10 p.m. the aircraft, having left this position, was sighted heading in a South Westerly direction, and some minutes beyond that it was seen again by ground observers near Coleford in Gloucestershire. Significantly, this time smoke was reported coming from the starboard outer engine. At this point the height of the aircraft was reported as approximately 500 feet.

When next seen the aircraft was heading in a North, North Easterly direction just 200 feet above the treetops passing over the roofs of the small village of English Bicknor, the observers on the ground all reported seeing large amounts of smoke and a fierce fire billowing from the No.4 engine. Halifax V9977 finally crashed into a field on the southerly side of Coppet Hill, just South of the village of Goodrich. Its final resting-place was some 130 yards north of the River Wye, the structure having finally failed after being burnt through by the fire in the wing and the starboard outer engine.

Eye Witness To The Crash

The direction the aircraft was heading at the time of the impact was roughly on course for a return to base, but it is far more likely that Pilot Officer Berrington was trying to attempt a crash landing on the flat ground just beyond the village of Goodrich, the other side of Coppet Hill, and approximately 3 miles from where V9977 crashed.

What had happened? Incredibly, considering the remote location of the crash site, the final moments of Halifax V9977 were actually witnessed - indeed, the falling aircraft only missed the eyewitness concerned by about 250 yards! Onslow Kirby was a farm worker who had rented and worked Green Farm on Coppet Hill from its owners. On 7 June 1942, he had been standing in the pasture field, just to the south of the farmhouse itself, some 150 yards from the edge of Coldwell Wood. The following is a reconstruction of the final moments of Halifax V9977 based on Kirby's account which was given to the Air Ministry Crash Investigation authorities a day after the event.

From his position in the field, some 250 yards from where the main fuselage finally fell, Kirby first heard and then saw the stricken aircraft approaching him from almost due South. He recalled the time as approximately 16.20 p.m. The aircraft was flying very low and coming directly towards him from the direction of the village of English Bicknor which was hidden from view by the high ground of Court Wood and Raven Cliff, South, South East of Coppet Hill. At first glance he noticed that there was a fierce fire burning on the right wing (starboard), but he could not recall details such as whether or not the propeller was actually turning on either No.3 or No.4 engine. Nor could he determine which of the two engines was the source of the fire because of the amount of flames and smoke.

Figure 10.09 - The trajectory of Halifax V9977 to the crash point (Courtesy of William Sleigh)

At the point that he first saw the aircraft, Kirby estimated its height above the ground as approximately 500 feet, but by the time it was above Court Wood and Raven Cliff he was quite specific that V9977 only cleared the treetops by 20 feet or so. As the aircraft cleared the trees and passed over the South bank of the River Wye its altitude had dropped to just 350 feet above the ground. The height of the ground at Raven Cliff is approximately 320 feet above the level of the ground where Kirby was standing, so his estimates can be considered as quite accurate.

As the aircraft passed over the mid-point of the River Wye the structural integrity of the Starboard outer main spar failed. This is the part of the aircraft that holds the wing to the main inner spar of the fuselage. The outboard mainplane disintegrated making the aircraft unstable and unable to maintain level flight. At this point the Halifax rolled over through 180° degrees, ending up practically on its back, at the same time it dropped almost vertically towards the ground falling in an arc down towards the field in which Mr. Kirby stood witnessing the entire event. Halifax V9977 struck the ground on the South side of Coppet Hill where the field rises quite steeply from the North bank of the River Wye. The main fuselage section hit the ground first 130 yards from the River just below Kirby's direct line of vision, which was slightly obscured by a low ridge. The severed piece of wing followed the crashing aeroplane falling short of the main fuselage by 30 yards.

Kirby was standing between 250 and 300 yards away from all this and described the moment as an immense bang followed by an enormous explosion and a subsequent fire. Running over to the crash site, Kirby's initial reaction showed little concern for his own safety but, upon seeing the complete destruction of the fuselage coupled with the huge fire that now engulfed the remains of V9977 it became obvious that he would need to go for help. When it was self evident that there was little he could do on his own he ran back to Green Farm to telephone the Fire Service.

In the fierce inferno that followed the crash much of what was left of the aircraft was consumed. There was absolutely no chance of survival for any of the eleven members of the flight and it is assumed that they were either killed instantly upon impact or died very shortly afterwards in the fire.

First News

At RAF Defford the aircraft was due to have returned at about 17.00 p.m. O'Kane, Dee, Curran, Trott and Lovell were waiting, along with the rest of the TRE staff and the TFU ground staff at the aerodrome for the flight to return. Lovell and O'Kane especially were beginning to get concerned when they had heard nothing from Berrington following his one solitary radio contact. O'Kane kept a diary at the time, and the following list of events are taken from his personal records:
17.00 Lovell rang control. Not landed.
17.30 Rang control. No news and no contact. No interest.
17.45 Rang control. No news. Rang a flight officers mess, control: Nobody about.
18.00 Rang G/C King [the CO of Defford]
18.10 W/C Horner [an officer on King's staff] rang up and wanted to know what was up. Told him.
19.00 Nothing done. 19.15 F/Lt. Reynolds [another on King's staff] had just left control having informed Group.
19.45 First news of crash.

So by mid evening on 7 June the first reports of the crash had started to come in and the various members of the TRE/TFU were organized into a small convey of vehicles, among whom were recovery teams, medical teams, security personnel and of course the remains of the scientific team who had to see what could be recovered of the H2S equipment. All now left the crash site to see what could be done. From O'Kane's diary again:
21.00 Left for Welsh Bicknor.
22.00 Finding the less important cargo [that is the magnetron].

That evening, the small procession of vehicles meandered their way through the twisting narrow lanes of the remote countryside around the River Wye. As darkness fell at the end of that summer's day in June, and with the light in the West barely illuminating the scene, the teams from RAF Defford came across what remained of Halifax V9977 now spread across a significant part of the pasture field in which it fell.

Recalling the events of that night in 1977, for the BBC TV series The Secret War, Lovell explained: "The reports of a crash in south Wales began to come in, and the rest of that night was just a nightmare. I was driven by the Commanding Officer of the aerodrome, a man called King, and winding through these lanes near Ross on Wye, searching for this wreckage and then the field with the burnt-out Halifax. And of course it was wartime. There was no time for emotions, our first duties were to search for the precious highly secret equipment, and collect the bits and pieces of it"

Very little had been spared by the force of the initial impact, and the huge fire that had followed. With hindsight it seems rather macabre that the bodies of those that died were ignored and took second priority to the retrieval of the radar equipment that the aircraft had been carrying but, such was the sensitivity of the work being carried out, that this was the instruction that the search and rescue team was told to work under.

Bernard Lovell, who of course would have been on the flight, and presumably would also have been killed, had he not given his place to an eager member of the EMI team, led the men searching through the wreckage. The Halifax had disintegrated on impact and the pieces of the aircraft, those that were left, were spread over quite an area - nearly an acre in fact. The importance of finding the magnetron and the rest of the H2S radar system was not lost on the team, and eventually they did recover all the pieces of the equipment.

As day gave way to night, a stillness returned to the field where the bomber had fallen. Armed guards were posted, the remains of the eleven who had perished were covered with sheets, and the teams went back to Defford with the shattered remnants of the precious radar equipment - King and Lovell arriving there around half past midnight. The rest, including the recovery of the bodies of the aircrew and the TRE/EMI team would have to be left to the next day.

Lovell would write in his diary the next day of the enormous tragedy he must have felt losing so many close friends and colleagues: "Monday, June 8 [1942]: Yesterday was terrible. Our Halifax V9977 crashed and killed Hensby, Vincent, Blumlein, Browne, Blythen, S/L Sansom, and a crew of 5 including the Pilot Berrington. All day Saturday (sic Sunday) we were having a meeting [at Defford] with EMI. I arrived at 2.30 to continue with Ryle etc.: they were just about to go off. They said they'd make it short. By 5 I began to worry. It took until 7.15 to stir up any interest in it. At 7.55 we heard it had crashed, at 8.35, 11 bodies were taken out, at 9, Group Captain King [the commanding officer at Defford] was driving O'Kane and myself down to salvage the apparatus. It was about 6 miles SW of Ross in the Wye valley. It was a mass of charred wreckage, quite unbelievable. We salvaged some bits but there wasn't much except the magnetron recognisable. Arrived back 12.30 at Defford and finally Malvern at 2"

The Air Ministry were informed by signal the next morning at 09.00 a.m. from RAF Madley that the flight had been lost with all on board, and it was organised for a team of crash investigators to visit the site the same day, Monday 8 June. The EMI team also went out to the crash site on that Monday morning though with all the security they had some difficulty at first getting anywhere near the field where the Halifax lay as Felix Trott recalls: "There was a guard at the entrance to the field where the aircraft had come down, and this fellow, who had probably been there throughout the night, was armed with a 'Tommy gun' and wouldn't let us pass despite our protestations that we were civilian scientists. Anyway, I suppose this chap was only doing his job, but you see he would not respect anybody that didn't have a uniform, and we'd driven over from Tewkesbury where we had stayed the night, without anybody from Defford to accompany us.

"Then, I suddenly remembered my military pass which I had with me, and decided it was worth a try showing him this. He looked at and decided we were probably OK, at any rate, he wasn't going to argue with authority as he saw it, and we entered the security area; but without that pass I don't suppose we would have been allowed to enter. The aeroplane had hit the ground inverted and had struck a small hump in the middle of this quite large field, exploding on impact. I could see that the undercarriage had been put down by the pilot, though now it was sticking up into the air from the remains of the engines. There were bits of aircraft everywhere, twisted and burnt out and none very large at all. On the far side of the crash area, down towards the river, the army or whoever it was, had erected a small white tent and in there, the bodies had been laid out. I didn't go in, I didn't want to see them, but I did catch a glimpse as the tent flaps were pulled aside and I saw far more than I wanted to"

Figure 10.10 - Felix Trott's military pass which was used to gain access to the crash site on Monday, 8 June 1942 (Courtesy of Felix Trott)

Eric White also recalled the horrific scene: "By this time there were people walking around all over the place. Some were military others were not. I expect that they were something to do with the crash investigation team. We walked over to this little white tent that had been erected and there were some people from TRE already there. Inside were these blackened, burnt bodies which were beyond identification and I left very quickly".

Among the people walking around that morning was an official photographer, though whether he was from Defford, attached to TRE, or part of the Accident Investigation Branch is not known for certain. What is known is that this man took at least six photographs of the crash site, which were then taken back to Defford for processing. They were then entered into an RAF log book of negatives as plate numbers 'G/59' to 'G/64' under the title 'Halifax Crash'. The plates were exposed on Wednesday, 10 June 1942, plate G/59 showing a 'general view', G/60 the 'port engine outer', G/61 the 'starboard outer engine', G/62 a 'field corner view', G/63 the 'fuselage' and G/64 the 'four engines'.

These photographs were logged as 'required by Wing Commander Horner', who was attached to TFU at Defford, and despite the log clearly stating that they were booked out and then returned, the plates have mysteriously disappeared. They are no longer to be found in the archives at Malvern. Indeed, it is quite possible that for purposes of secrecy, they were removed from the archive files and any prints made from the plates, even the plates themselves may have been destroyed. Other documents relating to the crash mention more photographs taken that day, in one instance 13 photographs were supposed to exist. Whether these are a completely separate set, or include the six mentioned above, is not known. What is certain is that the very highest authority in the land ensured that the secrecy surrounding the loss of Halifax V9977 was total. None of these photographs of the crash site has yet been found in any archive, and may not do so for many years to come, assuming that is, that they still exist at all.

Back at EMI that Monday morning, Maurice Harker arrived for work and happened to get into the elevator with a very subdued-looking G. E. Condliffe. Harker said 'good morning' to Condliffe who turned to him and replied, "I don't suppose you've heard yet then?". "Heard? Heard what?", and it was then that Maurice Harker was told that Blumlein, Browne and Blythen had all been killed the previous afternoon. "I simply could not believe it", he recalled, "I walked into my office and just sat there staring out of the window. It was such complete and utter disbelief. I said to Condliffe 'what do I do? Shall I just carry on as before? And he said yes, carry on as before. And of course, I had to collect what things I needed and drive back to Defford that afternoon which I did. But I didn't go out to the crash site once there. I don't suppose there was any need, and I certainly didn't want to go"

Three days after the crash, on 10 June 1942, Lovell would write to his wife Joyce in Swanage where she was then still living with their two young children, "The accident has left a great aching sore in our minds which is always there. We try to forget it by working hard and thinking that we just have to make a tremendous job of it to make up for that enormous sacrifice." And in his notes to himself he wrote, "The loss of Blumlein is a national disaster. God knows how much this will put back H2S?". Even two weeks later Lovell's diary thoughts were still immersed in the crash, "Last week we tried to forget about the crash and get on with the work. This week we have perhaps succeeded better"


Naturally the loss of the scientific personnel aboard V9977 was a crushing blow to British radar development, but in time of war there is precious little space for sentimentality. The development of H2S was continued even before the official crash investigation report was printed. Halifax V9977 was replaced with Halifax W7711 on Friday, 26 June 1942, and along with R9490 the two aircraft completed the trails of the magnetron and klystron systems.

As predicted by the EMI team that had died, the magnetron system proved far too good not to use as a production system, and when, on Wednesday, 10 February 1943, W7711 returned to active duty with the RAF it became the first aircraft that had been fitted with the H2S radar system for operational use.

When the war was over the government did allow a brief tribute to the men who had lost their lives throughout the conflict in the Telecommunications Flying Unit. But a full and fitting memorial was not erected until the dedication of the memorial window at Goodrich Castle on 7 June 1992, fittingly the fiftieth anniversary of the crash of V9977.

Figure 10.13 - Memorial Window at Goodrich Castle unveiled on 7 June 1992

In the aftermath of the loss of the aircraft and crew, and the subsequent investigation that followed, no blame was attached to any person or single factor as prime cause of the crash. True it was a degree of negligence that had resulted in the tappet lock nut not being fully tightened; the Rolls-Royce report clearly indicates they were aware of the implications of such an oversight. Having said that, no other Merlin engine suffered from this particular problem and no other documented case of this type of failure is known. It would seem that the loss of V9977 was due to a freak accident that was literally one in a million.

During the years following the crash of V9977 and long after the end of the war, speculation among the family and friends of the aircrew and the scientific team aboard the flight did not abate. With little if any factual evidence being made available to them beyond the obvious, that their loved ones had died in the service of their country, it was not surprising that many, including Doreen Blumlein, believed, wrongly, that the crash was the result of sabotage by the Germans.

The Crash Site Today

I also wanted to visit the site of the crash myself, if for no better reason than to try to understand the layout of the terrain and the harrowing experience that Onslow Kirby must have gone through that day in June 1942. The land on which V9977 fell is part of the Courtfield Estate, owned by Patrick Vaughan Esq., who kindly gave me permission to visit the crash site in the company of Douglas Kirby, Onslow Kirby's son. I would have liked, of course, to have visited the site with Onslow Kirby, but sadly, he had passed away some months before my visit, before I had had the chance to talk to him personally.

The roads around the villages of English Bicknor and Welsh Bicknor are windy and narrow, and it would be easy to get lost trying to follow a map to the small private road that leads down to Green Farm and Coppet Hill. It was in the high summer of 1996, when I made my visit to the crash site. On a Sunday, when the weather was much as it must have been on 7 June 1942, I stood in the exact spot where Onslow Kirby had witnessed the stricken aircraft fall. It was almost fifty-four years to the day since the crash of Halifax V9977. There is no monument or marker to indicate the location in the field itself, which is still farmed and regularly ploughed. Had it not been for the fact that I was looking for a place where an aircraft had crashed, the actual site could be said to blend perfectly into the rest of the countryside unnoticed. If you look more closely however, and if you have a guide who is aware of the significance of the place, then you can find the place that I had come to see.

There is a clearly visible indentation in the earth at the exact spot where the fuselage hit the ground which, despite the years that have passed, acts a poignant reminder of the events that took place there. Onslow Kirby kept a small plastic bag in which tiny fragments of V9977 were wrapped, brought up in the years that followed the crash during the ploughing of the field. There were twisted cartridge shell cases, a small fragment of Perspex and a twisted and burnt piece of metal, probably part of a strut. After his death, Kirby's plastic bag and its contents disappeared, and with it the last tangible memories of the crashed aircraft.

The path that V9977 took as it headed across the treetops towards Kirby, would have taken only few seconds of time to pass. His view of the River Wye as it meanders through the shallow valley, would have been obscured by the small ridge just away in front of him on the route to the riverbank. Beyond this, on the far bank, are the trees themselves which were cleared by just 20 feet. The treeline hides any view of the road or the village of English Bicknor, which lie just behind the ridge. At this point, the ridge forms part of a large, steep embankment on the south side of the river, and rises up to Court Wood and Raven Cliff. A narrow public footpath, part of the Wye Valley Walk, follows the southern bank of the River below the ridge, as it passes directly opposite the point where the Halifax fell. It is a path surprisingly little walked today, even in this picturesque part of Monmouthshire so often used by ramblers.

Figure 10.14 - The view from Goodrich Castle looking west towards the area just one mile further on
from the crash site, where Halifax V9977 might have landed safely

I found myself standing there in that field for quite some time, silently wandering what must have gone through the mind of Onslow Kirby faced with witnessing so cataclysmic a scene. Other than the knowledge that I possessed of the tragic events that had taken place on that day in June of 1942, it is fair to say that there is nothing of great distinction about this particular spot. Only that it is a quiet, beautifully serene, picture of English country life; timeless and peaceful, probably unchanged from that day to this - a day when the peace was so cruelly and suddenly shattered.

These much shortened excerpts have been taken from two chapters of the book
'The Inventor of Stereo - The Life and Works of Alan Dower Blumlein'
by Robert Charles Alexander, ISBN 0-240-51628-1 Focal Press, publ.


On Tuesday, 10 September 2002, a memorial will be unveiled by Sir Bernard Lovell, dedicated to the
RAF Aircrew, Scientists, Engineers and Civiliam Personnel
who lost their lives in the furtherance of Radar Research between 1941-1957.

Naturally, I shall be attending. Details of this memorial can be found using the link below:

Defford Memorial Unveiling 2002

Return To Main Menu

All material copyright ©1999, 2000, 2001, 2002, 2003, 2004, 2005,
2006, 2007, 2008, 2009, 2010, 2011, 2012
Robert Charles Alexander All Rights Reserved