From Nighthawk to Spitfire

by John K Shelton

  The following extract from G.A. Cozens’ manuscript ‘Concerning the Aircraft Industry in South Hampshire’ describes the humble beginnings (1913) of the firm at Oakbank Wharf, Woolston, which Reginald Mitchell joined three years later:

  Supermarine [as it was soon to be known] seems to have begun almost by accident and in the early stages the unpredictable nature of the firm’s founder [Pemberton-Billing] and his equally colourful general manager [Scott-Paine] might have diverted the destiny of Supermarine in any one of several directions. The factory was in a part of Mr Kemp’s boatyard just above the Floating Bridge on the Woolston side of the River Itchen, and a number of strange contrivances were built there. Mr Kemp often said that it was he who kept the little firm going, and indeed the works facilities like the sawmill were very useful and the workforce, who were largely the Kemp boatyard men at the start, were versatile and able to carry out some unusual projects.

  (G.A. Cozens lived close to these works in the early days of the company, and was a school friend of one of their workers as well as a neighbour of Henri Biard, their long-serving test pilot. He has left some fascinating, often anecdotal, information concerning the local aircraft industry, particularly on the subject of Supermarine. The author believes that this extract, and the others following, while sometimes inaccurate, deserve to be more widely known.)

  THE NIGHTHAWK AND OTHER

  EARLY AIRCRAFT

  Many accounts state that Mitchell joined Supermarine in 1917, but we find, in Sea Flyers, C.G. Grey writing that Mitchell ‘had been discovered by Mr Pemberton-Billing as a competent draughtsman and later Hubert Scott-Paine put him in charge of the design department’. Noel Pemberton-Billing was no longer the owner of the company when it became Supermarine in 1917, and the Flight and Aeroplane obituaries to Mitchell in 1937 both stated that he joined the company in 1916. Also, a surviving works drawing of the PB 31E Nighthawk, relating to the central nacelle, its gun mountings, and the various cable runs, is initialled ‘R.J.M.’ and dated 18 September 1916.

  The young Mitchell, builder of small model aeroplanes, must have felt that any misgivings about moving, alone and so far from his family and home, were forgotten, at least temporarily, when he arrived at Pemberton-Billing Ltd and saw this aircraft – a 60ft-span monster quadruplane, standing nearly 18ft high. It was completed and first flew (presumably with the control cables and gun mountings drawn by Mitchell) in February 1917, some months after Pemberton-Billing Ltd was bought by Hubert Scott-Paine, on 20 September 1916. It was he who adopted the old firm’s telegraphic address for the new firm – Supermarine Aviation Works Ltd.

  The first version of the Nighthawk machine was the PB 29E of 1916 which had been devised in response to the frighteningly new bombing raids by German airships. In order to reach and patrol at the heights attainable by these invaders, an aircraft with a large wing area was required. The usual biplane principle for lightness of construction was applied with a vengeance and resulted in its quadruplane configuration. It crashed not long after its delivery to naval pilots for handling trials but, soon afterwards, the young Mitchell was working on a new version of this ‘gunship’.

  Just as Mitchell’s career ended with, and was brought to the attention of the wider public by, the creation of a land-based fighting machine, the Spitfire, so his design career, devoted mainly to marine aircraft, started with a non-maritime fighting system. The 962sq. ft wing area of the PB 31 was designed to support the weight of a Lewis gun in the nose and, unusually, a non-recoil 1.5-pounder cannon mounted in the top wing pylon, together with another Lewis gun. Equally unusually, there was also a 5hp engine and generator installed in the fuselage to power a movable searchlight at the very front of the aircraft for the purpose of searching out airships at night.

  Under a well-known American test pilot of the day, Clifford B. Prodger, the new aircraft was found to reach 75mph and to have a landing speed of only 35mph, but it also took one hour to reach 10,000ft, as its two 100hp Anzani engines proved not to be powerful enough. Mitchell, therefore, must have learned swiftly how aircraft designs were, more than anything else, at the mercy of engine technology.

  Meanwhile, the problem of successfully combating the German Zeppelins had been solved by 1917 with the development of an explosive bullet that enabled sufficient oxygen into the airship’s hydrogen bags for them to be ignited by an incendiary bullet. As both these bullets could be loaded into the gun magazines of conventional aircraft, the development of the heavily armed Nighthawk was not required.

  An earlier Pemberton-Billing design, the PB 25, had fared better with an order for twenty machines, and these had a configuration that was to become very familiar to the later Southampton firm: the small biplane with a pusher engine. Like the PB 29 and PB 31 it was also a land plane, as the First World War mainly encouraged the development of aircraft to operate over the battlefields of France and Belgium.

  The embryo company gained further valuable design and structural information as their wartime effort involved repairing and building of other firms’ aircraft – in particular, twelve Short S38 and twenty-five Norman Thompson NT 2B seaplanes.

  PB 31 Nighthawk under construction (R.J. Mitchell fourth from right?). (Courtesy of Solent Sky Museum)

  Design experience of a more direct sort soon came Mitchell’s way for, when he joined the company, it was completing an order to build some Short Type 184 floatplanes. The association with this firm was particularly important as those firms granted licences to manufacture Short’s products were supplied with full sets of blueprints and had to send their staff to the parent company for instruction. Other budding aviation firms such as Fairey, Westland, English Electric and Parnall also benefited from this arrangement – not to mention the untrained Mitchell himself.

  The company had also just finished two flying boats specified by another pioneer aviation group, the Admiralty design team. This unit, from the Royal Naval Air Station at Eastchurch, together with the one at the Royal Aircraft Factory at Farnborough, represented much of the contemporary ability to tackle aerodynamic and structural problems in a scientific manner.

  This first group of people was responsible for designing for the war effort and as a result, they were instigators of the Handley Page heavy bomber and a range of Sopwith aircraft that were originally intended for naval use. Most of the early theoretical work, in particular the seminal Handbook of Stress Calculations, came from this source, and so it was again fortunate for the young Mitchell that some of this leading team was sent down to the works at Woolston to draft out the details of new naval machines.

  The first two aircraft completed to Admiralty designs were accordingly known as ‘AD Boats’ and again exemplified the pusher biplane configuration of the PB 25. With the addition of the PB type of wing superstructure to the new boat-like hull, there begins to appear the general flying boat formula that was to inform Mitchell’s Commercial Amphibian, Sea Eagle and Sheldrake, and which led up to his well-known Walrus. Significantly, the particular details of flying boat hull construction also came to the new company: F. Cowlin, the technical supervisor at the Royal Naval Air Station has recorded how he went down to the Pemberton-Billing firm and ‘learned a great deal about hull design from Linton Hope, who joined the section for a time while we were engaged on the AD Boats’. The lines and structure laid down by this well-known yacht designer became the basis of all the wooden flying boat structures that Mitchell subsequently utilised and were, in fact, a considerable advance on that of the current naval flying boat, the Curtiss-derived Felixstowe.

  Cozens also supplies relevant information on very early hull construction and workforce conditions:

  In 1914 the firm built the small flying boat PB I which was a credit to the workforce, and indeed it was judged to be the best example of aircraft construction at the 1914 Olympia Aero Show. This applied to the workmanship but unfortunately not to its performance.

  The PB I hull was of round construction, built by small boat methods with close
ly spaced wooden ribs of half inch square section like girl’s hoops, joined by longitudinal stringers and covered by two layers of mahogany or cedar wood planking, laid so that the outside layer was sloping the opposite way to the inner layer, this was known as ‘opposed diagonal planking’. There was a layer of doped fabric between the layers and the whole fastened by brass screws or copper nails and in some cases the nails were turned over and clinched or riveted over a small washer …

  The machine never flew and it is likely that the aerodynamics were wrong. The engine and propeller were at 15 degrees to the centre line of the hull, so that it seemed the machine was intended to lift off as soon as the engine opened up. In any case, the engine was not powerful enough to maintain flight.

  An Echo ‘Letterbox’ contributor wrote to say that his father, who was working at Supermarine, was told to fetch an axe and Pemberton-Billing, after looking at the beautiful machine for a long while, broke it up.

  When the Great War started, the Supermarine factory [not yet so named] became involved in repairing damaged floatplanes from Calshot and no doubt the workforce gained experience from this, but the PB 9 was a typical Supermarine venture. It was almost certainly a copy of the successful Sopwith Tabloid which had won the first [second] Schneider Trophy at Monaco the year before but there is no doubt that the plans for the PB 9 were little better than something on the back of an envelope, probably hastily drawn and given to a foreman, who was then told to do his best. However, the simplicity of the design and the construction did not detract from its performance and it handled well.

  Scott-Paine trusted his men to work well and he knew that they were the ultimate ones in whom he had to put his faith, it was long before the aircraft inspection board was established. The remarkable thing about the PB 9 was that it only took seven days from the time the project was set in motion until it was finished, and a number of stories have been put forward as to how this was done, but the way Scott-Paine worked was always the same, and instead of locking the men in until they had finished, as one story said, he agreed with the foreman on a price for the job and left it to him. No doubt the men did work hard and long, but in their own way. They took into account the fact that there was inadequate lighting in the factory, and worked as much as possible during daylight, and they also had in mind the fact that the copal varnish had to dry and other practical things, so that, whatever the urgency was, the practical considerations were what counted …

  The worst fear of craftsmen employed at Supermarine was that by some mischance his job could be lost for the most trivial accident, even breaking a thin twist drill meant a walk to the storekeeper to ask for a new one and the payment of a fine. The problem was that the flying boat hulls were fastened by rows of small screws which meant drilling through two layers of cedar ply and into the rock elm ribs, which was hard on drills, but before long the men solved the trouble by making their own drills. They cut knitting needles or piano wire into lengths of about 2in and annealed one end, hammered the tip flat and hardened it like a spear. Each man had an Archimedes drill with a simple brass chuck which could be purchased as a part of a fretwork set and was suitable for drilling small holes at awkward angles because it was light and only needed a straight [manual] push and pull action …

  Supermarine workers with tools of their trade. Mitchell is standing (and looking uncomfortable?) third from right. (Courtesy of P. Jarrett)

  By 1917, the first of twenty-four production AD machines to be ordered was undergoing acceptance trials for the navy, and so Mitchell, who obviously knew much more about heavy locomotives than lightweight wooden aircraft, had ample opportunity to see the flying boat construction techniques employed as the rest of the aircraft were built. As Harald Penrose of Westland wrote:

  The Admiralty had found floatplanes too dependent on smooth water; they were interested in the far heavier flying boat hull which in the Linton Hope approach consisted of a double skin of mahogany planking with fabric in between, with rock elm strips forming almost circular ribs, longitudinally stiffened by closely spaced stringers.

  However, the hydrodynamic aspects of aircraft hull design were in their infancy. The early test pilot, John Lankester Parker, described his first acquaintance with the AD Boat as follows:

  Not only did it develop a formidable porpoise at a very low speed, but nothing I could do would prevent it turning in ever smaller circles to the right, despite the fact that my passenger went out on the port wing-tip to keep one float well and truly in the water.

  Aside from structural matters, the young landlubber Mitchell would have now discovered that it was one thing to design an efficient yacht hull but quite another to produce one that would easily plane over the water and break free from its suction in a controllable manner. On the one hand, a calm sea and no wind might prevent the currently low-powered aircraft from being able to ‘unstick’; on the other, it might ‘porpoise’ in a series of bounces over waves until the right flying speed and angle might be achieved. (The standard First World War Felixstowe flying boat embodied an alternative type of hull with a flying boat fore body married to a conventional land plane fuselage of longerons and struts, but it was found to need additional planking, which affected its performance as well as its flying trim. It was no better a performer on water than the AD Boat.)

  The unlikely transformation of a lad from the middle of England trained in locomotive engineering into a designer of sea-based aircraft became possible because Scott-Paine must have recognised something about the applicant for the post of personal assistant when he stood before him – although his particular engineering background was not necessarily a drawback. For example, H. Fowler had been apprenticed to the Lancashire and Yorkshire Railway, and had become the chief engineer of the Midland Railway before becoming the superintendent of the Royal Aircraft Factory, and S.T.A. ‘Star’ Richards had been apprenticed to the Great Western Railway before becoming Handley Page’s personal assistant and, later, his chief designer in 1922. In those days, Mitchell’s mathematical and drafting skills would have been his predominant qualification, and applicants for Mitchell’s post would most probably have only offered aviation experience as an additional bonus.

  Whatever had singled out the young man from the Midlands for his first post, first impressions were clearly not ill-founded, for Mitchell was promoted to assistant works manager in 1918, and his improved financial position now enabled him to travel back to Stoke and to marry Florence Dayson (headmistress of Dresden Infants’ School) at St Peters church, Caverswall, on 22 July. It is a striking fact that Florence was eleven years older than himself, which suggests that the young Mitchell’s mentality was attracted to that of a mature professional woman and that he was displaying at an early age a single-mindedness and determination that overrode what, particularly at that time, must have seemed to his parents an unusual match.

  One must also imagine that some of Mitchell’s early domestic evenings were spent in study, as the same year saw the issue of HB 806 by the Technical Department of the Air Board which contained a full account of the mathematical methods employed by the department. And while a contemporary designer admitted that in such vital matters of weight/strength ratios and, therefore, safety margins, ‘we did it by guess and by God’, more theoretical information became available for Mitchell to study in 1919, with the publication of Aeroplane Structures by A.J. Pippard and J.L. Pritchard, and Applied Aerodynamics by L. Bairstow.

  The recent appointment of a professor of aerodynamics at Cambridge also marked the development of something approaching a systematic and scientific approach to the new technology. But how far the Southampton company was to become a significant part of it was by no means certain. The Society of British Aircraft Constructors did not have a committee member from Supermarine and the strong British presence at the 1919 Paris Air Show did not include any machines from Mitchell’s firm.

  However, at least the new assistant works manager was busy early in 1919 with another seaplane project, namely
the conversion of some of the surplus Admiralty AD Boats into civilian passenger-carrying aircraft. Ten of these two-seaters, which were in storage, were purchased back from the Admiralty with a view to offering trips from Southampton to various seaside resorts on the Isle of Wight, and drawings were prepared for the installation of four passenger seats and a more economical engine.

  The name ‘Channel’ indicates the very modest transport ambitions of the company and, in fact, the first of the passenger services was only between Southampton and Bournemouth. But at least Supermarine had the distinction of obtaining the first British Certificate of Air Worthiness for a passenger-carrying flying boat in the August of 1919. The flights, which cost 4 guineas single and 7 guineas return, were claimed to be the ‘First Flying Boat Passenger Service in the World’. (One of the pilots employed was Captain Henri Biard, who was to test all of Mitchell’s designs in the next ten years.)

  Several trips were also made to Cowes, but only for passengers who had missed the regular ferry. However, by the August of 1919, regular services to the Channel Islands as well as to the Isle of Wight were begun, weather permitting, in addition to joy rides at various venues on the south coast when opportunities arose. During that year, the steam-packets ceased operation in sympathy with striking British railway men and the Channels finally lived up to their name by operating a service to Le Havre for the duration of the dispute, so providing a precedent for the government’s use of the RAF to provide a newspaper service between London and the provinces during the 1926 General Strike. The service, costing £25 return, began on 28 September and came to an end on 5 October; the Bournemouth Service gradually petered out with the onset of winter.