Precursors of R.J. Mitchell's Walrus.

Seagull III, Papua (from painting by author).

The first success of Mitchell as Chief Designer was the Commercial Amphibian and this had brought an order from the Air Ministry for a  military development of the type as part of its policy to assist the struggling aviation companies to stay in business. This was obviously not a philanthropic gesture but a recognition that British air power needed the support of a healthy aviation industry especially as, by 1920, the RAF had been in action again with the new military tactic of ‘control without occupation’ – a very economical and swift-acting aerial alternative to the employment of large army land forces in the policing of colonial and League of Nations mandated territories – against tribesmen in Mesopotamia, Transjordania, the Sudan, and Somaliland, against the Bolsheviks in Russia, and against the Afghans on the Indian frontier.

SEAL II
In order to fulfill the Air Ministry requirement, Mitchell’s next design was to be a three-seat amphibian for use as a fleet spotter; it would be required to be extremely seaworthy and to have the lowest possible landing speed with good control – in order to land on to aircraft carriers. Mitchell’s response was known as the Seal II, presumably with the Commercial Amphibian being regarded as the Mark I predecessor.
     The Seal II had the outwardly-retracting landing wheel geometry first introduced on the Commercial Amphibian but the detail of the Seal’s retracting system shows that something had been learned from the criticisms of the earlier plane’s mechanism: the earlier machine had an undercarriage of two steel tubes, hinged below the lower centre-section join with the lower main planes and the wheels were raised or lowered by sideways pressure from a tube in the hull to the wheel axles; the new undercarriage now had a single strut, suspended from the lower wing and braced by two tubes hinged to points on the hull; it was moved inwards for retraction by means of a worm and bevel gear under the wing. This revised method was utilized on all future Supermarine amphibians up to and including the Sheldrake of 1927.
   Of the flying surfaces, only the tailplane followed the previous Hargreaves Sea Lion outlines and, although Mitchell placed the stabiliser lower on the fin, he retained the inverted aerofoil principle. This feature was also continued in aircraft up to the Sheldrake, being necessitated by the high thrust line of the engine which caused increasing nose-down forces as power increased. With the tailplane being ‘of the depressing kind’, as it was described in the publicity for Hargreaves’ earlier Sea King, the need for constant back pressures on the stick was therefore obviated at cruising speeds.
   On the other hand, the wing shape was new and this planform was retained by Mitchell for all his subsequent single-seat naval aircraft, up to and including the Sheldrake. But the rearward-folding wing requirement for a shipboard aircraft had not been tackled by the Supermarine company since the Baby of World War I and Mitchell adopted a similar approach – and one which he, again, continued with in military aircraft until the Sheldrake: the forward wing strut at the joint between the wing centre section and the main plane was doubled so that one of these members carried the weight of the leading section of the wing when folded back. The need for wing-folding also required large cut-outs to be made in the trailing edges of the wings, so that they could fold close to the plane’s centre-line to keep storage space to a minimum, and the wings were placed further forward than in the Commercial Amphibian so as not to project behind the trailing edges of the tail assembly when folded.
   The wing-tip floats, with their decreased side-area, were less clumsy than before and could be carried on struts to the waterline. The pilot was placed well forward and supplied with a machine gun which could be retracted and shielded during take-off and landing; the wireless operator was just aft of the wings and the rear gunner behind him but with the fuel tanks separating off the pilot from these other two crew members. 

    Because two of the crew members were placed behind the wings, a tractor layout had to be chosen for the engine to prevent the centre of gravity moving too far back. In Supermarine’s publicity for the Seal, attention is drawn to this placement, no doubt because of its novelty (at least in single-engined machines):

the engine is the Napier of 450 h.p. The engine mounting is unusual in that it is of the tractor type. This has been rendered possible by the fact that in this case the greater part of the useful load carried is aft of the wings in the tandem cockpits, and the success of the tractor mounting will allow this type of boat to be arranged either as a tractor in such a case as this or as a pusher in cases where the greatest useful load is concentrated forward. Very great attention has been paid in designing this engine installation to securing accessibility for inspection and adjustment of the engine and its accessories.

One notices the offer of the more conventional pusher layout – presumably in the hope of a civilian version which would not, one might reasonably assume, need a gunner behind the wings. The company publicity also draws attention to the new designer’s typical concern with the practicality of his machines – in this case, the ease of access to the engine.

As N146, the Seal II first flew in May, 1921 and, in the following year, one machine was sold to Japan who were keen to be kept abreast of Western technological developments. Despite this general lack of orders, the Seal is important in our story as it is the one early Mitchell design which most clearly looks forward to one of his three main aircraft types – the Seagull II to Seagull V/Walrus series of medium sized amphibians.

Meanwhile, the company had to come to terms with the post-war Anti-Waste League and the resultant Geddes Committee Report which led to a drastic reduction in all Government expenditure. The new Secretary of State for Air, Sir Samuel Hoare, reported that, in 1923, only 371 front line aircraft remained, either in the British Isles or abroad, and assessed the current situation thus: ‘Orders for military planes had almost come to an end and a demand for civil planes did not yet exist … Only two thousand five hundred men and women were left in the industry and the few firms engaged on machines and engines were on the verge of closing down.’
   On the other hand, the actions mentioned earlier of ‘control without occupation’ had to be backed up by support for the ailing aircraft industry. Thus it had been decided that, over the next five years, thirty-four new squadrons would be formed, bringing the air defence of Great Britain up to fifty-two squadrons by 1928. [In the event, the total home squadron numbers only rose to thirty-four by the date proposed.]

SEAGULLS II to IV
The first positive result of the new situation was seen when Supermarine received a letter from the Air Ministry which suggested that it ‘might be inexpedient’ to close down the works entirely as Supply and Research were considering an order, ‘the exact amount of which cannot yet be stated, but which might approach 18 machines, spread over the period ending March 31st, 1924’.
   In response to this Air Ministry lifeline, Mitchell modified the Seal II machine. A more powerful Napier Lion II engine was fitted – again in a tractor layout – and the fin consequently increased in area and the wingspan reduced. Thereafter, two new aircraft, N158 and N159, were completed by March, 1922, by which time, the wing-tip floats had been redesigned, the wings given a slight sweep back, the ailerons redesigned, and the fin area further enlarged. The number of modifications resulted in these first production Seagulls being designated Mark II.
   There was one particular modification of the Seal type which ought to be mentioned: the fuel tanks had now been moved from the fuselage to positions under the top wing centre-section, thus supplying petrol to the engine by gravity feed. [A Supermarine comment on the previous Sea King II design had been that ‘the petrol supply is by pressure, and every effort has been made to reduce the length of piping and eliminate as much as possible the carrying of piping into the hull’.] Perhaps because of the weight distribution of the basic Supermarine flying boat configuration, there appears to have been no undue problems created by the new fuel arrangements and there was also a particular bonus: Supermarine was now able to announce that ‘inter-communication between crew has been considered fully, and a through passage is arranged for this purpose’. [In passing, it should be noted that the constructional methods of the Linton Hope hull that Supermarine had adopted conferred another advantage as there were no internal bracings or structural bulkheads to be weakened by making a passageway between the pilot and the other crew members – see my Blogpost: ‘Mitchell’s Wooden Hulls, Structure and Finish’.]

A competitive test on HMS Argus between the Seagull and the Mark VII version of the Vickers Viking had found in favour of Mitchell’s machine and an  order for five Seagulls, N9562 to N9566, was then received in February 1923. This was followed by a requirement for another thirteen (N9642 to N9654), also in 1922. These aircraft equipped No. 440 (RAF) Fleet Reconnaissance Flight and some were placed aboard the aircraft carrier HMS Eagle. An additional machine was again sold to Japan.
    Further orders for the Seagull came in when the Australian government decided that their Air Force should assist in the hydrographic survey of the Great Barrier Reef. 101 (Fleet Cooperation) Flight was formed on 1 July 1925 and six Supermarine Seagull III amphibians were ordered (A9-1 to A9-6). These machines were essentially Mk IIs with larger radiators and the first of these was ready by February, 1926 – by which time six of the RAF aircraft had served a tour of duty with HMS Eagle but, thereafter, the type was pronounced as having ‘no potential naval use’.
  In sharp contrast, the Australian Seagulls were used more thoroughly, as their survey work extended into 1927 and continued on northwards to include some 10,000 square miles of Papua and one staged flight of 13,000 miles. Interestingly, three RAF Seagulls, engineless, had been acquired at the scrap price of one hundred pounds each and were intended to be used for spares. However, they were found to be in such excellent condition that they were re-erected and quickly put into service.
   This traditional Supermarine ruggedness was also evident after the survey work was completed, as the Seagulls were assigned to the newly constructed seaplane tender HMAS Albatross, commissioned in 1929, and continued in carrier use until 1933 when the vessel was placed in reserve.

The Seagulls had, by this time, lost their appeal for the RAF but, as a result, the Seagull design continued to occupy the minds of Mitchell’s design team even until 1928: fitting hydro-vanes was considered and, because of its porpoising tendencies, various permutations of the hull step position were tried out on N9565 and on N9606. And one aircraft, N9605, was fitted with Handley Page slots and a new tail unit with twin fins and rudders. This aircraft, designated Mark IV, was converted to take five passengers in 1929, when the Supermarine company was looking forward to the old Sea Eagle Southampton–Channel Islands routes being resumed [see my Blogpost: 'R.J. Mitchell's Sea Eagle Fleet'] – with a small fleet of Seagull IVs. A pilot service was begun in July by the prototype five-passenger conversion (G-AAIZ) but most of August was void owing to serious damage to the hull caused by its hitting a barely submerged rock. Then, on 2 September, the short-lived business ceased when the aircraft ran into engine trouble.
   Two other Seagulls, N9653 and N9654, were converted for civilian use: registered as G-EBXH and G-EBXI respectively they began a coastal service at Shoreham but this also failed, owing to inadequate public response.
   However, two other modifications of the Seagull were of great significance to Mitchell’s team. One was concerned with equipping a Seagull to initiate the testing of catapults for launching aircraft and the second was the changing of the usual water-cooled Napier engine to an air-cooled radial engine in a pusher configuration. As we shall see later, when the Seagull V/Walrus appeared, it was an aircraft engined in this particular way and stressed for catapult launching.

* * * * *

For reference sources, see my Blogpost: “Source Material and References" – an extended bibliography is included in my R.J.Mitchell at Supermarine; Schneider Trophy to Spitfire  which also provides material for wider reading, grouped according to specific areas of interest. 

R.J. Mitchell's Sea Eagle Fleet.

In 1922, the Air Ministry gave approval for an air service between Southampton, Cherbourg and Le Havre. The route, with a subsequent extension to the Channel Isles, was to be operated by an air service named the British Marine Air Navigation Company and Hubert Scott-Paine and James Bird of Supermarine were to be its directors.

The aircraft which Mitchell was to design for this service was to be powered by the Rolls-Royce Eagle IX engine, hence the chosen name – “Sea Eagle”. Despite the somewhat aspirational name, Mitchell went back to the more boat-like hull shape of the larger Channel and Commercial Amphibian designs and, in fact, the fore section of the Sea Eagle resembled a cabin cruiser of the time, with its high, pointed prow, enclosed accommodation for passengers, large windows and grab-rails running the length of the passenger compartment above the cabin and along the top of the hull. As the two planing steps were also joined by a continuous hard chine which ran three-quarters of the hull length, it embodied more than any other Mitchell design the original Pemberton Billing concept of ‘boats which fly’.

    Presumably because of a need to consider local hangarage for the aircraft, wing folding was again adopted. It would seem that width rather than length was the important consideration because a forward folding arrangement was again adopted, which also had the structural advantage of folding at the main spar, although this arrangement necessitated a cut-out in the leading edge of the wings, which did nothing for aerodynamic efficiency.

    Mitchell continued the practice of gravity feed for the engine of this latest flying boat with apparently little qualms about stability problems, for the fuel tank (and subsequently a second tank) was now attached to the top of the centre section. On 28 June, a Flight correspondent wrote that that ‘this machine represents a great step forward in the development of the seaworthy [commercial] amphibian’ having appreciated the ‘most important innovation’ that, in place of the usual tank in the hull, ‘the main petrol tank has been mounted on top of the top plane, so that direct gravity feed, with its attendant simplicity and freedom from breakdown, can be used’. The writer also added, ‘the fact that the engine is mounted high above and some distance aft of the cabin has resulted in reducing the noise audible in the cabin to a minimum, and as a matter of fact, in the “Sea Eagle” it is possible for the passengers to converse in ordinary tone of voice, without having to shout to one another’.

One departure from all previous (and future) practice was the use of a pronounced stagger of the two wings, as the weight of the forward passenger cabin and its six passengers necessitated bringing the centre of lift of the top wing well forward of the engine. (One suspects that Mitchell’s usual preference was for the simplicity of directly opposed biplane wings.)

Sea Eagle at Woolston

The first of the completed Sea Eagles had made its maiden flight in June 1923, and received its Certificate of Airworthiness on 11 July. Two days later, Supermarine entered the new aircraft in the King’s Cup Air Race that had been initiated by the Air Ministry the year before, also to encourage aviation development. As it was a handicapped event, the entry of a commercial flying boat might not seem too strange but the carrying of four passengers must have had much to do with the Company being mindful of publicity generated by air races. Unfortunately, circumstances involving a burst tyre and its replacement led to the aircraft being disqualified.

   On the 5th of the next month, the Director of Civil Aviation at the Air Ministry, Sir Sefton Brancker, came to Southampton and was given a display of the machine’s ability to negotiate the (usually) crowded seaway as well as a demonstration flight. He announced himself to be well satisfied with the Sea Eagle’s potential contribution to the development of civil aviation, both in terms of performance and comfort and, along with other senior members of his department, had another flight nine days later.

   Particular comment was made on the very sensible placing of the passengers and, in the following publicity, the Company makes reference to the advantages of this arrangement:

The passengers are accommodated in a roomy cabin in the fore part of the hull. This cabin is very comfortably fitted out. Its position in front of the engine makes it very quiet and free from engine exhaust, gases, oil, etc. . . It is very efficiently heated and ventilated, and is fitted with sliding triplex windows along the two sides for use in the warm weather.

One passenger recorded descending into the Sea Eagle and finding ‘a delightful little room’ that the Company had fitted with ‘reposeful armchairs’.

Regular daily services between Southampton and Guernsey began on 25 September, 1923, so constituting the very first British scheduled flying boat service; it was advertised to leave Woolston at 11.15 a.m. and return from St Peter Port at 3.30 p.m. (The French section of the service did not materialise.) The service, often with breaks due to bad weather, continued with the Sea Eagles for the next five years, even though the single fare to the Channel Isles was not cheap. Compared with boat transport, however, the normal flight time of one and a half hours was very attractive although, in adverse wind conditions, it might be almost an hour more.

   Nearly four years later, the fleet of three Sea Eagles was down to one: G-EBFK having crashed on May 21, 1924 (reportedly due to a bird strike) and G-EBGS was rammed and sunk when moored at St Peter Port on the 10th of January, 1927 – a reward of £10 for the identity of the culprit was never claimed.

In the event, Mitchell’s machines had not only operated the first scheduled flying boat service in Britain but they also had the distinction of forming part of the basic fleet of the organisation which eventually became British Airways – on 31 March, 1924, Imperial Airways Ltd was incorporated as the 'chosen instrument' of the British Government for developing national commercial air transport on an economic basis and the British Marine Air Navigation Co. was one of four companies taken over for this purpose. The two Sea Eagles which were remaining by that time now had their fuselages painted with prominent ‘Imperial Airways’ lettering and they continued their accustomed service to the Channel Isles under the control of the new national company from May 1, 1924.

The last of the three Sea Eagles, G-EBGR, was finally retired in 1928, thus justifying Supermarine’s claims that this type was ‘very strongly built and very seaworthy’. A photograph from a correspondent to The Aeroplane showed a Sea Eagle hull at Heston Airport in 1954. It had been presented to BOAC in September 1949 and intended for restoration and display at the new London Airport; but nothing came of this proposal and this piece of industrial archaeology was burnt on 13 February, 1954.

* * * * *

For reference sources, see my Blog: “Source Material and References. " An extended bibliography is included in my R.J.Mitchell at Supermarine; Schneider Trophy to Spitfire  which also provides material for wider reading, grouped according to specific areas of interest. 

Mitchell's First Schneider Racers – the Sea Lions II and III.

Mitchell's Sea Lions had a common ‘ancestry’ in the N.1B Baby design and the Sea Lion I had a limited input by Mitchell (see my Blog:"R. J. Mitchell's Early Modifications"), whereas the quiote different Sea Lion II was entirely his responsibility – as, of course, was the Sea Lion III which followed.

Sea Lion II

In 1922, the Air Ministry issued Specification N6/22 for a single-seat fighter capable of operations from aircraft carriers or as a floatplane. The winner of the contract was the Fairey Flycatcher – it had a slightly higher top speed than Mitchell's Sea King (133 compared with 125 mph), it was just as aerobatic, its short span allowed it to be struck down to the hangers without folding wing arrangements, and its extensive aileron-cum-flap arrangements produced very low minimum take-off and landing speeds.

     With hindsight, it might thus be seen that the days of a fighter flying-boat were numbered but Scott-Paine, the managing director of Supermarine, was still determined to continue with the type and decided to seek publicity for it by entering a version of the aircraft in the 1922 Schneider Trophy contest. Another, more patriotic, reason might have been to prevent the Italians from winning the Trophy outright with a third win in the forthcoming competition (after the inconclusive Bournemouth event of 1919, the Italians had had fly-overs in the following two years).

     Also, the rules had been changed in 1920 to encourage a more practical type of aircraft rather than an out-and-out racer: 300 kg. of ballast had had to be carried and, although this requirement was dropped the next year, it was replaced by a watertightness test in which the aircraft had to remain afloat fully loaded for six hours. These rules tended to suggest the flying-boat’s suitability for the Trophy contest and it must have been noted that the recent Italian designs to meet these requirements had, indeed, been flying-boats.

Despite the omens favouring flying-boats, the uncertain financial outlook of the Supermarine company at this time was such that its Managing Director did not feel able to incur the cost of designing and building an entirely new machine and so the fuselage of the previous Sea King II was utilized. He also obtained the loan of a Lion engine from the manufacturers, a high speed propeller, petrol and oil from other manufacturers, and a fifty percent reduction in insurance rates.

     Mitchell, no doubt remembering the sleek Savoia at Bournemouth, aimed for increased speed by making the entry of the fuselage, originally shaped to house a gun, somewhat smoother (see also "R.J. Mitchell's Early Modifications"). And as the Lion engine to be fitted developed 150 hp more than the original Hispano-Suiza engine, he was able to decrease the area of the wings by reducing their width. Another modification was necessitated by the test pilot's refusing to test fly it until the rear fuselage had been stiffened up. Again, in response to the extra power of the engine, an additional increase in fin area was called for. Mitchell achieved this with least expenditure of time and money by merely modifying the vertical surfaces above the tailplane; the leading edge of the fin was given a pronounced forward curvature which proved to be effective but certainly won no prizes for elegance.

     The finished machine was named "Sea Lion", thus drawing attention to the name of the loaned engine. It was also designated a Mark II, to distinguish it from Hargreaves’ earlier design but, as a result, misleadingly suggested that it was a direct development of it, contrary to the pedigree described above.

Supermarine Sea Lion II  (from a painting by the author)

As there have been many accounts of the Supermarine 1922 win, I will confine myself to the following relatively unknown contemporary Cozens extract:

. . . bad weather cut down Captain Biard’s chances of getting used to the Sea Lion, and this was further jeopardised by a forced landing which began with the engine cutting out over the Dock. However, when he had had a few more flights he was satisfied and the speed and handling proved very good, indeed it was faster than any flying boat or seaplane of that time. Then, with the limited time available, it was doubtful if they could get the Sea Lion to Naples in time but the General Steam Navigation Co. agreed to take it and it was hurriedly dismantled, put into a crate and on to a lighter, and one of Ray's tugs took it down to the Solent and the freighter Philomel lifted it on board and took it to Naples.

Pre-race spying and counter-spying on both sides was all part of the event. This atmosphere continued throughout the whole series and it was the policy of each competitor to arrive at the start of a competition with a machine that was ahead of its rivals by virtue of some secret and outstanding advantage which was not revealed until it was too late for anyone to copy. In the case of the Sea Lion this meant that the wing-span was cut down to an absolute minimum and as the trials at Woolston had been curtailed even Captain Biard was not too well practised as to the machine’s behaviour.

He kept is speed down in the practice flights but was quietly getting used to the course and conditions, and his engine fitter, Mr. Pickett tuned up the engine to the higher temperature of the Bay of Naples, so that when the race started Biard was reasonably prepared …

When Captain Biard and his victorious team came to the Floating Bridge with the great prize held above their heads no-one bothered whether it was a Cup or a Trophy – everyone called it a cup, certainly Scott-Paine. I had parked my bicycle outside the Woolston Picture House and I saw the Supermarine workers run down to meet them. They had taken the two swivel chairs from the office and fixed them to poles and they lifted Captain Biard and Scott-Paine in the chairs shoulder high and carried them round the works … Fireworks were let off and there was some horn blowing … [One suspects that Mitchell’s relatively recent arrival in the firm at the time as well as his temperamental self-effacingness account for there being no mention of him in the celebrations.]

Mitchell’s machine not only won the Schneider Trophy race for Britain at an average speed of 145.7 mph but also gaining the first F.A.I. World Records for seaplanes:

(i)        Duration – 1 hr 34 min 51.6 sec

(ii)       Distance flown – 230 miles

(iii)      Fastest time for 100 km closed circuit – 28 min 41.4 sec (130 mph)

(iv)      Fastest time for 200 km closed circuit – 57 min 37.4 sec (129.4 mph)

 Sea Lion III

As the next Schneider Trophy contest was to be held in England, it was to be expected that Supermarine would be only too happy to capitalise on their 1922 publicity by competing, successfully it was hoped, without the cost of overseas travel and accommodation. Indeed, the new venue decided on was to be Cowes, less than 20 miles from the Supermarine works at Woolston. But only with the large-scale production of the Southampton, which began in 1925, might Supermarine have felt justified in the cost of designing and building a one-off specialist racer and, as the top speed of the Sea Lion II was significantly less than that of the record breaking Savoia S.51, he did not immediately respond to the challenge.

    When Scott-Paine was persuaded to submit an entry, he confined himself to asking his Chief Designer to do his best with the 1922 airframe, to which Mitchell fitted an uprated Napier Lion III engine and radiator into a more streamlined nacelle. The more powerful engine also allowed for a reduction in the wing span by four feet and he also had fairings made behind the two hull steps:

Sea Lion III

In addition, he designed new wing-tip floats offering less frontal area, mounted them on streamlined struts, and added fairings around the main strut attachment points. Because an extra 75 hp was available from the new Lion engine, the rudder and fin were increased in area, the resultant combination looking less improvised than that of the Sea Lion II.

Mitchell’s changes could hardly prevent the Supermarine entry from showing its, by now, venerable pedigree and one feels that the rather whimsical sea lion head depicted on the nose and floats of the Supermarine entry was almost a self-deprecating gesture in face of the expected serious opposition from America: their Curtiss racers were fitted with of one of the great aero engines in aviation history, the Curtiss D-12, whose frontal area was about 50% less than the rival Napier Lion and their winning of the 1922 Pulitzer race was also due to the incorporation of radiators flush-mounted on the wings and to the use of metal propellers (as tip speeds were now approaching the speed of sound.) 

Sea Lion III at the competition base.

Of the three machines which finished the competition, these Curtiss CR-3 floatplanes came first, with an average speed of 177.3 mph, and second, with an average of 173.46 mph. Despite 75 more horsepower, Biard could only manage an average of 157.17, a poor third.

     Clearly, the usual European flying-boat formula with an engine mounted above the hull was no longer likely to be the best approach – apart from the formidable in-line engine and the flush fitted radiators, the Curtiss CR-3 configuration also limited the number of other drag inducing items to 16 struts, with 20 wires, despite the extra penalty of floats, whereas the Sea King/Sea Lion I tradition Mitchell had inherited had required 33 struts and 42 wires as well as a boat-like hull. Afterwards, Scott-Paine praised the Napier engine ‘that would have gone on for ever’ but said that he needed ‘to apologise to Capt. Biard because we did not give him a good enough machine’.

The Sea Lion, which had retained its Mark II registration G-EBAH, was returned as N170 to the Marine Aircraft Experimental Establishment at Felixstowe with its undercarriage and sea rudder/skid now restored but its career was short lived, owing to its extremely lively take-off performance. As Biard had said, ‘It was an interesting sensation; you switched on the engine, and before you could count 1, 2, 3, 4 fast – she was flying.’ Unfortunately, when Flg Off. E. E. Paull-Smith, took over the Sea Lion, he apparently did not take sufficient notice of the warning that the machine tended to lift off before flying speed had been reached. As a result, he took off, dropped back onto the water, rose to about forty feet, stalled again, and dived in. Paul-Smith was killed and the machine was too extensively damaged to be considered worth repairing.

     This incident, on 25 June, 1924, marked the end of Supermarine’s attempts to interest the Air Ministry in the seaplane scout concept and also the beginning of Mitchell’s search for a worthy and dedicated Schneider Trophy competitor (see my Blog: R.J. Mitchell's Annus Mirabilis, 1925: Part II).

* * * * *

For reference sources, see my Blog: “Source Material and References. " An extended bibliography is included in my R.J.Mitchell at Supermarine; Schneider Trophy to Spitfire (see sidebar) which also provides material for wider reading, grouped according to specific areas of interest. 

Mitchell’s Amphibian Bomber and an Enigma; The Scarab and the Sheldrake.

Supermarine Sheldrake

Supermarine Scarabs

(from paintings by the author)

The Scarab, which first flew on 21 May, 1924, was a powerful fighting machine for its period and was the only Mitchell design sold exclusively outside the British Empire;  the other Supermarine amphibian at this time, the Sheldrake, was an enigma. 

 Sheldrake

The result of the criticisms of the Supermarine Seagull [see my Blogpost: 'Precursors of R.J. Mitchell's Walrus'] was that, in 1923, an Air Ministry order was placed for an improved version. The resultant aircraft was the Sheldrake whose flying surfaces were virtually identical to those of the Seagull but which had a more efficient boat-like hull very similar to that of the Sea Eagle . It is surprising that the noisy stabilising screens on either side of the engine were still retained, as was the separation of the pilot from the rest of the crew – which already had been addressed in the Seagull II [see my Blogpost: 'Pre cursors of R.J. Mitchell's Walrus'].

   Even more surprising was the apparent inactivity around the Sheldrake – perhaps the first flights of the Sea Eagle and of the Sea Lion III, the first production batch of the Seagull IIs, and consideration of the larger flying-boats, the Scylla and the Swan, were important factors here – but, for whatever reasons, its only known public appearance is recorded as late as 1927, at the Hamble Air Pageant, where it could be seen to be an obsolescent type.

 Scarab

Meanwhile, a year after the Sheldrake was ordered, the second of our aircraft made its maiden flight – the first of a Spanish order for twelve aircraft. There was no prototype for this aircraft – as it was in a great many respects similar to the Sheldrake, the Air Ministry order for the Sheldrake provided most of the design work.

King Alfonso of Spain was a regular visitor to the Hendon RAF Air Shows and must have had an early appreciation of the new British ‘control without occupation’ tactics whereby air power, not ground forces, was used to control insurgent tribes in areas of the British Empire or in areas mandated to Britain by the League of Nations. As a result, the Spanish Royal Naval Air Service asked Supermarine for an amphibian to be produced, capable of carrying a bomb load of 454 kilos, as it had been noticed the passenger carrying capacity of the Sea Eagle promised a suitable basis for a design.

   In the new, Scarab, version, the engine was returned to the more familiar pusher configuration as the crew were all now, more conveniently, grouped together in front of the wings – with the navigator/bomb-aimer also having a position in the hull immediately behind and below his cockpit position.  The fuel tanks were now removed from the hull and placed above the middle section of the top wing. The space not now required in the fuselage was used for twelve 50 lb. bombs which could be dropped via a sealable aperture in the bottom of the hull. Four 100 lb. bombs were also carried under the wings and the total weight of bombs carried amounted to the equivalent of six men. The Scarab also had a crew of three and had to carry a machine-gun, ammunition and a considerable amount of fuel – thus making it an attractive powerful fighting machine for its period, given its single engine.

   The first Scarab made its maiden flight on 21 May, 1924, but whether all the twelve that were ordered actually saw service is unclear. One was damaged on acceptance trials when its Spanish pilot hit the side of a Union Castle liner when taking off; and they had to survive a severe Bay of Biscay storm stowed under tarpaulins as deck cargo. Nevertheless, Scarabs were seen above Barcelona at the 1925 Royal Review of the Spanish forces by King Alfonso and they equipped a seaplane carrier, the Dédalo, a converted merchant vessel – being lowered into the water or raised from it by crane.
    They were based at Carageno from whence they took part in actions against Riff and Jibala insurgents in the Spanish Moroccan campaign, including bombing raids in support of an amphibian  landing at Al Hoceima.
    The Moroccan conflict ended soon afterwards in 1926 and so Supermarine publicity ran as follows: "A large number of these machines have been bought by the Spanish Government, and these have been in operation for the past year in Morocco with the most satisfactory results [my italics]."

No Scarabs were ordered by the British Government but the Spanish order for the twelve machines was, at that time, quite substantial, especially coming soon after the order for 25 Seagull II aircraft and the three Sea Eagles.

* * * * *

For reference sources, see my Blogpost: “Source Material and References" – an extended bibliography is included in my R.J.Mitchell at Supermarine; Schneider Trophy to Spitfire  which also provides material for wider reading, grouped according to specific areas of interest. 

R.J. Mitchell's Ugly Ducklings: The Scylla and the Swan

During World War I, Felixstowe F 2 and 3 flying-boats had been operated successfully on coastal-reconnaissance duties, despite their various problems,  and they were replaced after the armistice by the F.5 from the same makers. However, D’Arcy Greig (later to figure in the Supermarine S.5 story) recorded that: 

they were grossly underpowered by two Rolls-Royce Eagle VIII engines, and if there happened to be a flat calm at time of take-off, they frequently refused to unstick. On such occasions the pilot had to taxi frenziedly up and down the Solent and around in circles in order to disturb the surface of the water before trying again, but even then they sometimes failed to get airborne.

These flying-boats had slab-sided hulls which were prone to leakage and so the Air Ministry was concerned to see if the Linton-Hope type of hull (see my Blog: "R. J. Mitchell’s Wooden Hulls – structure and finish"), which they regarded as a success on the early Supermarine seaplanes, could be adopted on aircraft of the Felixstowe size.

SCYLLA

The slow process of addressing this concern by various manufacturers around the country lasted well after the war, by which time Supermarine came into contention in 1921 with the Scylla, a triplane with biplane stabilizers. It was to be powered by two main engines and a much smaller, auxiliary one, sited in the hull to drive a water propeller for taxiing [and perhaps to assist with take-offs?]. This last feature seems particularly old-fashioned, looking back to the earlier Pemberton Billing days of the company, in particular, the P.B.7, and to a proposed Torpedo Carrier triplane of 1921.

Why only the hull of the Scylla was completed is unknown, as was its final fate. Most likely, Mitchell’s rapidly developing confidence as a designer was an important factor: as a new contract was received soon afterwards (Spec. 21/22) for a large commercial amphibian, his thoughts could turn from the traditional thinking represented by the Scylla to a more forward-looking aircraft. Also, required to carry twelve passengers, it would come out at about the same size as the military Scylla and therefore it might be that the first machine was soon relegated to merely providing information for the new design – which, as a biplane, Mitchell must have predicted would be a more efficient aircraft; it could also surely be easily retro-fitted for such military purposes as might have been envisaged for the Scylla. Certainly the former machine seems to have been used only for water taxiing trials and a photograph in Supermarine Aircraft shows a very basic framework erected on the hull to accommodate (temporarily?) the two engines for this purpose.

SWAN

When the new machine, to be named the Swan, first appeared in 1924, it was a considerable re-design, being an equal-span biplane with a forward-folding wing arrangement like the Sea Eagle of the previous year. Indeed the Swan might, in some ways, be regarded, rather, as a scaled-up Sea Eagle although doubling the number of passengers to be carried necessitated accommodating them in the main body of the hull rather than in the fore position of the earlier aircraft. Again, the fuel tanks were placed high enough to provide gravity feed to the engines which were situated between the wings, as well as to provide unusually roomy and fume-free accommodation for the passengers. The fin and rudder outlines also resembled those of the Sea Eagle.

     On the other hand, Mitchell’s less complex use of dihedral only on the outer sections of the lower mainplane was new and the three vertical tail surfaces anticipated his larger designs of the next decade. The single plane stabiliser was also new to larger Supermarine aircraft and was kept well clear of the water by the upward slope of the rear section of the hull – very unusual for the time although not quite an innovation in flying-boat design (see the much smaller French Tellier T3 or Latham HB3); nor was it as graceful as the upward sweep of the future Southampton rear fuselage but, at least, it represented a bold new step in Mitchell hull design, without there being previous experimentation with smaller hulls.

Supermarine Swan

The need to mount three fins had also led to a reversal of his earlier practice, whereby the tailplane had been supported by the fin (with the aid of numerous struts). The new feature also anticipated most of Mitchell’s later seaplanes and was a more elegant configuration than the traditional biplane-tail approach proposed for the Scylla. On the other hand, the upswept fuselage and the boat-like prow, flared outwards at the top to counteract spray, were features in common with the earlier machine which had presumably been proven to be effective by its taxiing trials.

     The raised cockpit superstructure was also very reminiscent of the unfinished Scylla and it contributed significantly to the clumsiness of the hull profile. Cozens’ comments:

The Swan had several features which showed improvements on previous designs, and no doubt these led to its success. The keel had an upward curve towards the tail that enabled it to take off more readily and this feature was noticeable in all later flying boat hulls built throughout the flying boat era, even to the Saunders-Roe Princess of the nineteen fifties, and it is very apparent if one compares the pictures of the Swan with that of the [Supermarine] Channel I. The struts of the Swan’s centre section formed large W’s which made for great strength and the large fins and rudders and the considerable spacing between the wings made this aircraft a success from a handling point of view. At any rate, Captain Biard was pleased and so was the Air Ministry, but no-one could say that the Swan was a handsome machine with its rounded bow and strange looking cabin and the pilot’s cockpit at the top.

The “strange looking cabin”, which housed a crew of two, sat on the top of the main fuselage so as not to interfere with the passenger space and, as the proposed passenger windows had yet not been fitted, the offending side-view was unrelieved. The same had been true of the Scylla and, whilst Supermarine had no doubt chosen the latter’s name to suit its proposed military role, it might seem to others that the name reflected its appearance: according to Ovid, the beautiful Scylla was turned into a thing of terror and in Homer Odysseus manages to sail past her but not before she catches and devours six of his men. As the new design was to have a more pacific role, the new aircraft was named ‘Swan” although, despite its size, “ugly duckling” comes more to mind. 

The Swan nearing completion.

Be that as it may, the new aircraft was first flown by Biard on 25 March, 1924 and, at this time, displayed the triangular cut-outs in the leading edges of the wings to enable them to fold forwards. The Swan also had the sort of retracting undercarriage arrangement that Mitchell had designed for his single-engined amphibians but the much increased size of the new machine necessitated the novelty of some form of servo assistance. Biard described the mechanism as follows:

it would have been quite impossible to wind down the six-foot wheels and powerful landing-carriage, which had to stand the weight of several tons of aircraft and passengers! So a neat device was fitted to the machine to do the work quickly and efficiently for us. This consisted of a small propeller, which, when not in use, was set sideways to the direction in which we were flying. When we wanted to lower the landing-gear, this propeller was swung round to face the direction of our course, and the whirling propeller was connected by cogs to a handle which wound very rapidly round and lowered the wheels into place; by turning the propeller rearward the wheels were wound up out of our way under the wings, and the machine was then able to descend on water. This gear, after one or two adjustments following minor troubles during tests, when the Swan behaved neither like fish, flesh, nor fowl, proved remarkably efficient, and wound the heavy landing-gear into place in about half a minute or less.

Biard also describes the visit of the Prince of Wales to Supermarine, and to the Swan in particular, on the 27th of June in the same year. :

An informal moment in the Royal visit, showing the impressive size of the Swan.

(Note the substantial samson post at the prow. Mitchell to right of Prince of Wales).

By the time of the Prince’s visit, the Swan’s two 360 hp Rolls-Royce Eagle engines had been replaced by Napier Lion engines, each developing 90 hp more than the Rolls-Royce units, which increased the Swan’s top speed by 13 mph; and the folding of the wings and the leading edge cut-outs had also been dispensed with. Plans for an RAF version of the Swan were also being actively pursued at this time – which may also have had some influence on the change to the fixed-wing layout and possibly throws more light upon the decision to terminate the development of the military Scylla.

Successful trials at the Marine Aircraft Experimental Establishment, Felixstowe, followed which were to have important results for the fortunes of Supermarine (see my Blogpost: “Annus Mirabilis, Pt.1) and, on its return to Supermarine, it was now fitted out for its passenger-carrying role. Company publicity pointed out that the machine was not only the first twin-engined commercial flying-boat/amphibian but that the provision of accommodation set new standards:

This is the first twin-engined amphibian flying-boat to be built in the world and it may also be fairly claimed to be the first twin-engined commercial flying-boat.

An important feature of this machine is that the whole of the hull is devoted to passenger accommodation. There are no internal obstructions of any kind, and the amount of room in the saloon far exceeds that of any commercial landplane. The internal accommodation consists of one large passenger saloon, elaborately furnished and upholstered and with every comfort. Forward of the saloon is the luggage compartment, fitted with racks for the stowage of passenger baggage. Aft of the saloon is the buffet, with all necessary fittings to supply light refreshment during the journey. Still further aft are the lavatories, which are efficiently and fully equipped.

As such it was registered as G-EBJY and first flew on June the 9th, 1926, carrying a representative of the newly formed Imperial Airways and eight excited female employees of Supermarine. A slight reduction in passenger seating had further allowed Supermarine to address the new standards in passenger accommodation, which The Aeroplane fully confirmed:  ‘the appointments are exquisite’ with ‘a commodious passenger saloon padded luxuriously and in which there are ten cosy armchairs. An ample porthole is provided for each chair.’

Mitchell (front left) in the Swan.

(Note the Linton-Hope fuselage structure of close-spaced hoops on longitudinal stringers)

The aircraft was loaned by the Air Ministry to Imperial Airways in order to supplement the service of Supermarine’s remaining two Sea Eagles, on their Channel Isles service. It operated during 1926 and 1927 but, as The Guernsey Evening Press reported, ‘during the normal rigorous inspection prior to leaving Southampton on April 12, a structural defect was discovered which necessitated the stripping of the whole machine’. As a result, the Swan was scrapped; Imperial Airways’ next long-distance seaplane was not to be the Swan, however, and so its main significance remained that of providing the prototype for the Royal Air Force’s next standard maritime reconnaissance aircraft, the far more attractive Supermarine Southampton of 1925.

* * * * *

For reference sources, see my Blog: “Source Material and References. " An extended bibliography is included in my R.J.Mitchell at Supermarine; Schneider Trophy to Spitfire which also provides material for wider reading, grouped according to specific areas of interest. More information, photographs and a three-view drawing of the Swan, as well as a full account of all Mitchell's completed designs and of the man behind them, will be available in a few weeks' time.