S.Ransom, R.Fairclough English Electric Aircraft and their Predecessors (Putnam)

Phoenix P.5 Cork

   On 31 October, 1917, a memorandum, signed by F.E.T. Hewlett on behalf of the Controller, Technical Department (TIN Section), was sent to Maj John Buchanan, Controller of the Air Board's Technical Department. The memorandum requested the Air Board's approval for a contract to be placed with the Phoenix Dynamo Manufacturing Co Ltd, of Bradford, for the design and construction of the aero-structure of two experimental F.3 flying-boats, whose hulls were then being built at May, Harden & May Ltd's Kingston Bridge Works at Hampton Wick on the River Thames. Approval was given under British Requisition No.256 and Sir William Weir, Controller of Aeronautical Supplies, was notified of the Board's decision on 12 November. Three days later, Phoenix was instructed to proceed not only with the design of the aero-structure but also with preparations for the production of the complete aircraft. Drawings of the hulls were forwarded and Phoenix was particularly requested to:
   '1) Increase the aerodynamic efficiency of the present machine [the standard F.3], keeping the wing roots exactly the same as the present machine in order that the aerial structure should be interchangeable with standard design.
   2) Simplify the design of the machine in order to facilitate production.'
   The contract, No. AS 37016/17, was signed on 28 November and the two flying-boats were allocated serial numbers N86 and N87. Phoenix gave the flying-boats the designation P.5, the appellation Cork not being used until about June 1918 after the issue of Technical Department Instruction 538 (later to become AP 547) regarding aircraft nomenclature. The Cork came in the category of aircraft with more than one engine and of 11,000-20,000 lb all-up weight to be named after seaboard towns in Scotland and Ireland.
   The P.5 was the result of an Admiralty decision to build a large twin-engine flying-boat to meet specification N.3B (later known as RAF Spec XXX), which the Admiralty's Air Department had formulated early in 1917, for an anti-submarine patrol machine of similar displacement to the Porte-designed Felixstowe flying-boats. The Admiralty decided also that the P.5 should have a hull designed by Lt Cdr Linton C. Hope, and authorised May, Harden & May to build two hulls to his principle of monocoque construction. This type of structure proved to be remarkably strong, yet resilient, and enabled the hull to have good streamline form. Generally, the Linton Hope hull was covered by two thin layers of narrow planks, the inner layer usually being wrapped diagonally and the outer laid longitudinally, rivetted to closely-pitched hoops attached to numerous stringers, which were in turn fastened to almost circular hoop frames, the whole being built up on a keel that ran from stem to stern. The planning bottom was added separately and was similarly constructed.
   Manufacture of the hulls, which were not identical, started towards the end of 1917. The first hull appears to have been completed shortly before 11 April, 1918, when Maj A.J. Miley, Assistant Controller, Technical Department (Design), informed Phoenix of the measured weight and position of the centre of gravity of the hull, and the second a little later. The first could be distinguished from the second by its planing surface, which, although discontinuous at the main step, extended aft to the rear step: the second hull had only a forward planing bottom and small rear step faired to the basic hull, the chine-line of the rear step being formed by a strake that was continuous with the top of the fin deck of the forward surface. (In R & M No.461: The Resistance of Flying Boat Hulls, published in July 1918, the first hull was numbered P.5 and the second, P.5A. This nomenclature was also used, though not widely, by Phoenix who applied the designations to N86 and N87 respectively. The order in which the hulls were eventually to be used however, was reversed, the P.5 hull being fitted to N87 and vice versa). The basic structure of both hulls was identical and each had openings in the upper surface for the bow gunner, pilot and copilot, who were seated in tandem, and the engineer. There were also two gunners' waist-hatches closed by detachable doors which had glazed portholes, although Linton Hope had, at one stage, considered a single cockpit on top of the hull for the rear gunner. Holes were also provided in the top-sides of each hull for the spars of the lower-wing centre-section. The hull's covering consisted of two layers of mahogany planks separated by a layer of varnished fabric for water proofing. The outer planks were laid longitudinally and were 5/32 in thick and the inner were wrapped diagonally and were 5/16 in thick, the whole being rivetted with copper nails and roves to 3/8 in square spruce timbers or hoops spaced 1 3/4 in apart. These had their ends let into the rock elm keel and were glued to 35 stringers of 1 1/2 by 1/2 in spruce, distributed evenly around the periphery of the hull and attached to almost circular frames placed at intervals of 2-3 ft. The frames were fastened directly to the top of the keel and were steam-bent from a length of rectangular-sectioned rock elm, joints being scarfed and made alternately at the top and bottom of the hull. The lower part of the hull was reinforced by short transverse strips of square-sectioned spruce, known as floors, which passed through the keel. Each hull's forward planing bottom had two mahogany skins separated by varnished fabric, the inner diagonal planking being 3/32 in thick and the outer 3/16 in, and fastened to a structure similar to that of the basic hull but with plywood transverse bulkheads, which created a number of watertight compartments. The rear planing bottom of the first hull was built in a similar manner, the thickness of its inner and outer planking being 5/64 in and 5/32 in respectively. That fitted to the second hull was little more than a fairing to the rear step but was also of Linton Hope construction.
   The first hull was delivered by road from Hampton Wick to Bradford before the end of April 1918, and was followed soon afterwards by the second, its journey northwards beginning on Wednesday, 8 May. Total production time for the two hulls was recorded as 31,348 man-hours.
   Meanwhile, on 9 April, May, Harden & May had delivered models of the hulls, which they had made to contract No. 38A/167/C.160 at a cost of ?26-8-0 each, to the Technical Department, where they were received by Linton Hope himself. These models were then made over to the NPL and used for comparative wind-tunnel tests, the results obtained being published in R & M No.461. The tests revealed that the first or P.5 hull had considerably less drag at all speeds than the second or P.5A hull. A modified P.5 hull, in which the fin deck was faired to the basic hull by a concave fillet, was also tested. It too was found to have less drag than the P.5A and was even marginally better than the original form. Despite the P.5's superiority, which would have been reflected in the performance of the complete flying-boat, it was stored on arrival at Bradford and the P.5A hull used in its place. The reason for this change remains obscure but possibly it had its origin as early as November 1916, when R & M No.300: Experiments with Models of Seaplane Floats, Eleventh Series was published. This report gave details of hydrodynamic tests conducted at the NPL on a model hull, similar to the P.5, supplied by Linton Hope. Various alterations were made to the hull and the form which was finally evolved was similar to the P.5A. This hull had better seaworthiness characteristics in that side waves created by the forward planing bottom finished as fan-shaped sprays clear of the stern. Previously, heavy spray was thrown up well above the top of the hull by the forward surface and the side waves tended to curl in and over the rear portion of the hull, with consequent saturation of and damage to wings and tail surfaces if they were set too low. It would appear that the P.5 hull was passed over in favour of the P.5A because of the former's poor seaworthiness. That the P.5 hull was built suggests that comparative full-scale trials were to be performed.
   At Bradford, in the spring of 1918, the aero-structure, which had been designed by W.O. Manning, was nearing completion. Its design had been done under the supervision of the Technical Department, which had checked and approved Manning's calculations and had made recommendations where necessary. An instance of this had occurred on 11 March, 1918, when Solly Brandt of the Structures & Stability Section had reported to Cdr Alec Ogilvie, Controller, Technical Department (Design), on the strength of wing spars, interplane-struts and bracing-wires, which, with the exception of a few control wires considered under strength, were found satisfactory. In accordance with Phoenix policy, that had resulted in the book Women on munitions of war being written by P.J. Pybus, Phoenix's managing director, at the instigation of the Prime Minister Lloyd George, all parts of the flight structure were designed to enable female labour to be used wherever possible. This policy was applied to the manufacture of wood and metal components. Of the latter fabricated items were reduced to a minimum, others were designed with mass-production in mind, and, in general, fittings and bolt sizes were standardised so that most parts were readily interchangeable and the variety of stores could be kept to a minimum.
   The unequal-span wings were made in sections: the upper wing in three, of which the outboard were braced from faired kingposts above the outermost interplane-struts, and the lower in four, of which two formed the lower centre-section, since on N86 the spars passed through the hull. Although a similar arrangement had been envisaged for N87, this section was built in one piece and attached directly to the top of the hull, the joint incorporating a streamline fairing. Each wing section had two box-spars made entirely of spruce and consisted of booms separated by distance pieces and webs joined to them by glue and screws. To reduce wastage the webs could be made from a number of short lengths spliced together, the number of splices, however, was strictly controlled. Ribs of RAF 14 aerofoil section were, in general, of Warren-girder construction and built from strips of spruce but those at the wing joints were fret-sawn from plywood and were lightened by elongated holes cut in them. Three-ply birch was used to cover the lower-wing leading-edge, to protect it from spray damage, and also the lower centre-section walkway used mainly by the engineer to start and tend the engines. Internal wing-bracing was by means of wires and tubular-steel struts, the ends of the latter being located solely by sockets in the spar fittings. Interplane and engine-struts were made of thin-walled steel tubes which had their ends sweated into and rivetted to socket attachments and which were faired by a fabric-covered wooden framework with three-ply leading edge. Reinforcing pieces of walnut or mahogany were fitted to the spars in way of connections for all struts and external bracing-wires, which were also of streamline section. Both wings were covered with fabric laid diagonally in broad strips and stitched to the ribs in the normal manner. Ailerons were fitted to the upper-wing only: each aileron was made in two parts to cater for wing deflections and was hinged to a secondary spar placed at a short distance behind the rear spars. The wingtip floats of N86 were identical to those fitted to the Felixstowe F.3; N87's were a little deeper.
   The tailplane was of similar construction to the wings but had an inverted RAF 15 aerofoil section and was made in two parts joined on the centre-line each half being braced to the hull by four faired tubular-steel struts. Three-ply was used instead of spruce for the spars' webs otherwise the materials used were the same as for the wings. Tailplane incidence could be varied on the ground by relocating the two bolts that joined the rear spar and passed through the rudder post. A horn-balanced rudder of composite wood and steel tube construction and of characteristic shape was fitted. The fin was readily detachable for covering and transport, having a diagonal box-spar which passed through the hull top and was joined to the keel by one bolt.
   Two 360 hp Rolls-Royce Eagle VIIIs were specified for and installed in each flying-boat, the engine mountings being similar to those used on the Felixstowe machines. Both engines were fully enclosed in cowlings made of flat wooden-backed aluminium panels. Three main fuel tanks were situated beneath the centre-section: the forward and transverse tank contained 80 Imp gal of petrol and the others 120 Imp gal each. All tanks were of the same diameter, and baffles, sumps and pipe connections were made alike. Two wind-driven Rotoplunge pumps delivered the petrol to two 20 Imp gal gravity tanks carried below the upper centre-section. The petrol system was duplicated so that either or both engines could be run from either gravity tank. The engine throttle control was the subject of British patent No.122,996 filed by Manning on 17 July, 1918. It consisted of a single lever connected by short cranks to two pulleys around which were wrapped the control wires to the engines. When the lever was pushed straight forward both throttles opened simultaneously and if it was moved to the right, the port engine was given more throttle than the starboard and vice versa.
   The basic armament of both flying-boats was four 230 lb or two 520 lb bombs carried below the lower wing roots and five Lewis guns, one at the bows on a Scarff ring, one each for pilot and co-pilot on either side of their cockpits on rail mountings, and one on a gallows mounting at each waist hatch. N87 had in addition two small nacelles fitted to the trailing-edge of the upper wing, each equipped with one Scarff ring-mounted Lewis gun and to which the gunners gained access by climbing the steps on the innermost rear interplane-struts. In August 1918, Phoenix were asked to study the possibility of carrying 520 lb bombs sunk into the hull end of the lower wing in the event that the bomb load as planned came too near the water. Comments were also requested as to how four 520 lb bombs could be carried. These proposals came to nothing, however.
   On 5 July, 1918, the Technical Department was notified of the completion of N86 and was asked for permission to erect the flying-boat at Brough, on the Humber Estuary. Exactly one week later another letter was sent stating that 'the P.5 [would be] ready for despatch very shortly'. Permission was given the following day and soon afterwards N86 was taken by road to Brough. Some parts, including the tail unit various struts and engine cowlings, were delivered by rail, however. These were sent on 26 and 27 July. Before N86's completion, Phoenix had asked that Clifford B. Prodger be allowed to pilot the machine on its initial trials, to which the Technical Department had consented on 4 July with the proviso that Maj M.E.A. Wright would undertake subsequent test flights.
   The first flight of N86, designated Cork Mk I, was made on Sunday, 4 August, 1918, between 6.24 and 6.33 pm by Prodger with Lt Cdr Hume observer, and Capt Slater, engineer, both of whom represented the Admiralty. Two more flights were made before 8 o'clock that day, one of 8 minutes and the other of 28 minutes. All took place in fine weather with a Force 1 northeasterly, wind blowing. Prodger reported that performance was very satisfactory. He thought, however, that N86 was slightly nose heavy and that the rudder was too small. On account of the latter he did not attempt to fly at maximum speed. After the trials N86 was moored overnight on the mud of the estuary foreshore and brought in at 5.30 the following morning. Between then and the next test flight made on the evening of 9 August, the covering of N86's wings was found to have stretched slightly but this factor apparently made little difference to the test as it was reported to be a 'fine performance'. A new and enlarged rudder had been fitted for this flight which lasted 54 minutes and was made with N86 fully loaded. Prodger was again at the controls and his crew comprised Hume, Slater and Finch, an engineer from Rolls-Royce. The test was witnessed by Maj Maurice Wright and Maj Miley who were to be pilot and passenger respectively for the allocation flight to be made the following morning. This test began at 11 :08 after some difficulty in starting the engines and was of 12 minutes duration. Because of the engine trouble another flight was arranged for the afternoon but it again proved to be a lengthy process starting the engines. By 4 o'clock, however, everything was in order but the tide was low. Nevertheless, a decision was made to proceed with the test. Unfortunately, just as N86 reached planing speed she struck a rock and started sinking, eventually grounding when half full of water. The crew worked hard to restart the engines, which had stalled soon after the accident, succeeded in doing so, and were able to taxi the flying-boat shorewards to ground her on the mud. By that time it was 7.30 pm. Salvage operations could not be undertaken next day because Phoenix-built F.3 N4416 was launched and test flown leaving no time to attend to N86. The following morning, however, the hull was patched and the machine brought ashore. After repairs, N86 was flown by Maj Wright to the Marine Aircraft Experimental Station, Isle of Grain, on 24 August, 1918.
   N86 passed final AID inspection at Grain the following day and during the ensuing week was test flown to obtain preliminary information of handling, climbing and speed, which was officially reported in NM 219, dated 31 August, 1918. Among the structural defects noted in the report was the slackness of the wing covering. Criticisms were made of the cockpit layout with regards to seats, instruments and throttle control, which was not considered as good as two separate levers capable of being grasped by one hand, this form being more familiar and natural to pilots. The criticisms, however, appear unjustified since, shortly after the hulls were delivered to Bradford, the cockpits were mocked-up using one of the hulls and their arrangement officially approved. The report continued: 'The machine is light on the lateral control and there appears to be sufficient [control]. It comes off a bank well. The rudder might be a little larger with slightly more balance and a higher gearing. The elevator is fairly light and effective. The machine is a little nose heavy. The boat is very stable in getting off and landing, no tendency to porpoise being noticed. A fair amount of water is thrown on the propellers when taxi-ing, and the bombs are washed by water thrown up by the fin (the outermost surface of the forward planing bottom), but not by a solid wave....The machine was taken off and landed in a slight lop. Apart from spray on the propellers the get-off was good, and the landing was very soft and free from. shocks'.
   An investigation into the cause of the slackness of the wing covering revealed that the type of dope used was unsuitable. As N87 was then under construction and its wings had been treated with a different dope, a decision was made to use them to replace those of N86. The exchange was effected in October 1918 and the original wings were sent back to Bradford for recovering and fitting to N87.
   It was evident from the preliminary trials that the wings of N86 were set too low on the hull. This fact almost certainly led to the wings of N87 being mounted on top of the hull (the holes in the hull for the spars were plugged).
   Trials with N86 were resumed in October after the new wings had been fitted. On 17 October, Grain Test Report NM 219B was published giving comparisons between N86 and the Felixstowe F.5 prototype, N90, of the view from the pilots' cockpits, layout of instruments, crew accommodation, handling, and take-off and landing performance. With the exception of instrument arrangement and crew accommodation, which were regarded as inferior in the Cork, there was little to choose between the P.5 and F.5. The author of the report might have mentioned (although in fairness to him he was probably not asked to record) that the P.5 was superior to the F.5 under the same operational conditions, in speed, climb and range, and in ability to carry a heavier payload. The report also noted that the P.5 had not yet been tried in rough seas and that the main step had given way once, possibly owing to its being strained on the machine's launching trolley. There followed another report, No. NF 2346, dated 22 October, concerning the rudder of N86 and this apparently led to the fitting of one on the lines recommended in NM 219. Overload trials were conducted and reported in NM 240b dated 10 May, 1920. At a later date N86 was fitted with navigation lights.
   N86 was thereafter used experimentally at Grain and was known to have been flying as late as April 1924, when it was the subject of impact tests described in R & M No.926.
   Construction and assembly of N87, which became known as the Cork Mk II, started about the end of August 1918. On 24 September, Phoenix were notified by the Technical Department of major alterations to be made to the hull. Fortunately by that time only the uncovered lower centresection and fin had been fitted (whether they had to be dismantled to allow the hull to be modified remains unknown). The alterations were made to Air Board drawing No. D.S.3.034: P.5 proposed production hull lines and were done at Bradford by May, Harden & May's boat-builders between 15 October and 2 November. The hull then resembled that fitted to N86 but had its rear step placed farther aft.
   From then, N87's assembly proceeded steadily and by 25 November the uncowered centre-section minus engines and the complete tail unit had been installed. However, further delay occurred when it was discovered that the wings would have to be recovered as the wrong type of dope had been supplied and used to treat them. The Cork Mk II was eventually delivered by road to Brough on 21 February, 1919, for final erection.
   N87's first flight took place during the late afternoon of Friday, 28 March, and lasted 7 minutes only, being curtailed because of strong winds. The crew comprised Majors Miley and Wright, pilots, CPO Dryden, Shanate of Rolls-Royce, West of Phoenix and Edwards, engineer. Gales and rough seas foiled attempts, in which a tailplane stay-tube was damaged, to bring N87 ashore and she was left moored to a buoy until the following Sunday evening when it was found that she had not shipped water. The Cork was observed to ride out the gales well and this fact was subsequently widely publicised by Phoenix, who compared N87 to its namesake. Despite the bad weather conditions the test was successful: the take-off was automatic, handling and stability, according to Maj Wright, was perfect, and N87 climbed strongly at 80 mph against a 25 mph head wind. Criticism was made, however, of the petrol pumps and cocks, the former for not delivering fuel at a sufficient rate, and the latter for being dangerously placed between the chain connecting the pump and its windmill. Recommendations were made concerning serviceability, in particular, for a towing-eye positioned at the stern for the connection of a rope, to minimise damage to the tailplane stay-tubes by manhandling. Undoubtedly, more test flights were made although these appear to have been interrupted when the wings were recovered yet again, Phoenix receiving a contract, No.PB35A/166/C.66, dated 24 May, 1919, for this work. N87 was flown by Maj Wright to the Isle of Grain in June and from then was used experimentally.
   In December 1919 the planing bottom of N87 was badly damaged when the machine was brought ashore on its trolley in rough weather. Repairs were made and completed early in 1920. One result of this accident was the invention of a trolley using inflatable water bags to cushion the hull.
   During the summer of 1922 a decision was made to install two 450 hp Napier Lions in N87. A trial erection of the engines, which were uncowled and mounted on a revised arrangement of struts, was made at Bradford. Afterwards, the engines and structure were delivered to the Isle of Grain and fitted to N87, which was then redesignated Cork Mk III. The first flight of N87 with its new engines lasted 10 minutes and was made by Flt Lt G.E. Livock on 2 August, 1922, the day it was to have joined the Flying-Boat Development Flight at the start of its cruise from Grain to the Scilly Isles. N87's flight had been delayed by a faulty petrol system. Flt Lt (later Group-Captain, DFC, AFC, RAF) Livock recalled the occasion saying: 'We had a rather 'do-it-yourself' system, constructed, I think at Grain and not by the makers, and fuel poured into the hull instead of the engines when we turned on the petrol. We had to do quite a bit of pumping out and mopping up before the flight'. N87 Joined the Development Flight cruise off Spithead on 4 August.
   The primary object of the cruise was to gain experience of operating flying-boats away from their home station using a ship as a base. The unit consisted of Short N.3 Cromarty N120, one standard Felixstowe F.5, N4038, one F.5 with Lion engines, N4839, and N87, accompanied by HMS Ark Royal, the parent ship, HMS Tintagel, a destroyer, and the RAF Floating Dock towed by HMS St Martin, a tug. Pilots assigned to the flying-boats were Flt Lt B.C.H. Cross, Flg Off Carey, Flg Off E.P. Davis and Flt Lt G.E. Livock respectively, and the Development Flight was commanded by Sqn Ldr R.B. Maycock. The final report and recommendations of the Flight were given to the Commanding Officer, Marine & Armament Experimental Establishment (Home), Isle of Grain, by Sqn Ldr Maycock on 31 October, 1922. The report began by stating the objects of the cruise in detail and listing the equipment and personnel involved. It continued:
   'Summary narrative of the cruise.
   Sheerness. The experimental flying boats were handed over to the flight on 13 July, 1922, and, with the assistance of a party of men loaned from M & AEE (H), the programme of work which it was considered essential to carry out to render them fit for the cruise, was completed, as far as possible by 31 July, 1922.
   All the flying boats, with the exception of P.5, which was only completed on the day of departure, were tested and found satisfactory before departing.
   The Ark Royal arrived off Grain on 25 July, 1922, and commenced taking in stores and petrol.
   Spithead. With the exception of N.3, which was left behind at the last moment to change an engine, the Unit sailed for Spithead and assembled there on 4 August, anchoring about 3 1/2 miles SW of Calshot.
   The Unit remained here for four days during which refueling experiments with the Ark Royal plant were carried out. This, at first gave trouble as a considerable quantity of water was found to be present in the petrol tank, the presence whereof has, so far, remained unexplained.
   N.3, N120, joined the Unit on 7 August, having had a new Condor engine fitted at Grain.
   Portland. The anchorage at Portland was very safe and no amount of wind at this base would have affected the security of the aircraft. It was, however, very congested and considerable judgement was necessary in handling the flying boats when coming up to and leaving the moorings.
   A considerable amount of engine trouble was experienced here with the Napier Lions installed in F.5, N4839, and Rolls-Royce Condors in N.3, N120, and minor hull and rigging defects were corrected in all flying boats. It was not possible, therefore, to carry out many concerted tests at this port and, moreover, it was considered by the Officer Commanding Unit that it was only on passage to its main base at Plymouth prior to leaving for the scene of operations and that it was not the intention to carry out trials at this stage. The weather at Portland was alternately good and bad.
   Plymouth. The Unit assembled at Plymouth on 14 August, 1922, and anchored in.
   A programme of exercises and tests was prepared, but little of this programme was actually carried out owing to a recurrence of engine failures and administrative difficulties arising out of the 'serious differences' reported by the Captain to have arisen between himself and the senior RAF Officer. Up to this period, however, a great deal of useful information and data had been collected on the general scheme of working with a Parent Ship, the maintenance and handling of flying boats as a unit, on the water, moorings, refuelling, and the inestimable value of the Seaplane Dock.
   The weather at Plymouth was similar to that experienced at Portland, wet and fine days alternating, and the effect of this was beginning to show up on the flying boats. Rust and corrosion could not be kept under and the planes were becoming groggy.
   Scillies. The Unit sailed for and arrived at St Mary's Roads, Scilly Isles, on 21 August, 1922. This anchorage possesses practically no shelter from the SW except for a few rocks, and the Atlantic swell very rarely, at this time of the year, ceases to roll in from that direction and, if accompanied by any wind, the sea becomes dangerous to any craft of small tonnage. In this respect this anchorage proved the most severe test of the seaworthiness of the flying boats of almost any possible to select.
   The south westerly swell was so continuous that it was found impossible to use the Seaplane Dock on more than an average of one day in four, owing to its excessive movement and the wash and send of the sea inside when submerged for docking.
   On days suitable for air operations, it was, generally speaking, quite calm, and the swell at the anchorage coupled with the fact that there was no wind, made the difficulty of getting the flying boats off into the air one of some hazard, and imposed a great strain on the hulls and superstructure....
   [The] Scillies is the most exciting testing place but, when working there, the best anchorage should be selected. This, however, was not done, as, in the first instance, the flying boats were moored out to sea and windward of the Parent Ship and the Destroyer. Later, they were moved to a more sheltered position but not until they had rode out one gale during which it was impossible to get them in and, to illustrate this, it may be mentioned that the ships' boats were hoisted or sent to St Mary's for shelter.
   At this base the maintenance of hulls and superstructure absorbed all the energies of the Flight and as it was not always possible to use the Dock, repairs were executed on a sandy beach and the pool off this beach provided the best anchorage for flying boats.
   The period spent at the Scilly Isles was the most instructive of the whole cruise from an experimental point of view and, although the numbers of hours flown was negligible compared with what could have been carried out had the flying boats been of a standard type, the experience gained in the proving and establishment of certain definite characteristics of flying boats - both affecting their possibilities and limitations - should prove of assistance in determining their future military value and to designers in evolving a seakeeping aircraft.
   [On 31 August, 1922, Flt Lt Livock made an aerial reconnaissance of Falmouth in conjunction with N4038, during which take-off tests were conducted in Mount's Bay and N87 was refuelled from HMS Tintagel off Penzance.]
   Eighteen days were spent at the Scillies and the weather prevailing was similar to that usually experienced during November. The flying boats had become so saturated and engines so corroded with salt water - due partially to their age, partially to the exposed anchorage and partially to the constant bad weather - that it was deemed advisable as a result of these conditions coupled with the loss of N.3, N120, from the same causes, to move the base back to Portland.
   Portland. The Unit sailed for and assembled at Portland on 8 September. N.3, N120, having been lost at the Scillies and N87 having sustained serious damage to the forward step during the flight to Portland, only the two F.5 types remained effective and with these, two successful air operations were carried out from this base, also some further experimental trials.
   On the completion of these operations and trials the Unit was prepared for the return flight to the Home Base at Grain, and, on 18 September 1922, the Unit sailed.
   Grain. The two F.5s reached Grain without incident but P.5, N87, came down at Newhaven, the outer aluminium exhaust manifold on the port engine having fallen off in the air and remained suspended by a wire stay.
   The Ark Royal arrived at the Nore on 19 September and came up harbour on the following day and, by 09.00 hours next day, all RAF personnel and stores were landed as the vessel was required for service elsewhere. '
   The report proceeded to give more detailed accounts of the performance of each flying-boat and the rest of the unit. Comparisons of the types were made and it was concluded that N87 was better than the Cromarty and far superior to either of the F.5s. The conclusion took into consideration each machine's condition at the start of the cruise and that of N87 was rated the worst! Appended to the report were the comments of the pilots on the performance and defects of their machines. Those of Flt Lt Livock read:
   General characteristics. No figures were taken for performance except take off which averaged 24 secs in a flat calm with no wind and full load up.
   The air speed indicator did not work correctly so it is difficult to gauge the exact full speed which is somewhere in the region of 101-104 mph.... The best cruising speed was about 78 mph, the leading speed [sic] being about the same as on F.5.
   Control. The controllability in the air is excellent with the exception of the aileron control which appears to be a little insufficient, this is noticeable at very low speeds such as stall landings and take offs.
   The rudder and elevator controls are very good.
   This type is much more pleasant to fly for long periods than the F.5, and is not nearly so tiring on the pilot.
   There is a very pronounced 'snatching' on the elevator controls when engine speeds are varied in the air. This may be due to the tail oscillation [caused by lateral bending of the hull].
   Control on water. The machine is very steerable on the water at low speeds, but sea anchors are required for picking up moorings etc, except in very strong winds, as the machine travels very fast through the water even with the engines throttled right down.
   The controllability at high speeds such as when taking off is good and the tendency to porpoise can be controlled to a great extent by the elevator.
   The difficulty in taking off is the danger of dipping a wing tip into the water. The wing tip floats being so low, any tendency of the machine to roll laterally has to be quickly checked, otherwise damage to the wing tip can easily result. This makes a cross wind take off especially difficult. When taxying across a cross wind at any speed it is quite easy to entirely submerge a wing tip float and have water running 'green' over the lower plane. Apart from this the machine is very clean when taxying, not making nearly so much fuss as on F.5.
   Seaworthiness. The seaworthiness of this machine was well tried out as six gales were ridden out without any damage whatever, and high seas, especially at the Scillies, were frequently experienced. It is interesting to note that one gale of Force 6, lasting nearly 24 hours was ridden out safely without the rear side doors in place. A small amount of water was taken in but not enough to endanger the machine. A certain amount of water, however, is taken in at these doors when taxying side to wind or turning quickly on the water.
   Loss of performance. The loss of performance, both in the air and taking off, as well as controllability was very marked towards the end of the cruise, this I consider due to the following causes:-
   (1) Hull becoming sodden with water.
   (2) Considerable leaking into the outer skin, which could not be kept under by the bilge pump.
   (3) Fabric on planes becoming slack. It is pointed out that the fabric was old and practically rotten before the cruise started.
   (4) Planes becoming out of truth from various causes.
   To sum up it is considered that this type [P.5] with the few small alterations recommended in this report should prove infinitely superior to the present F.5 in almost all respects.'
   Group-Capt Livock later recounted the last days of N87. 'The last flight [from Portland on 18 September, 1922] was very unpleasant. I just managed to haul her off the water after taxying for miles. In the air she was horrible on account of the enormous load of water carried in the step compartment. Off Newhaven, I had to force land owing to an exhaust manifold breaking off and directing the exhaust flames on to the magneto. I attempted to take off again after effecting temporary repairs but owing to her poor condition and having regard for the bumps felt from the crests of the waves this seemed inadvisable, so I taxied her into Newhaven harbor and tied up to a mooring in the entrance. That night a gale sprang up, the mooring dragged or parted and N87 crashed into a pier. The machine was then towed up the harbour to a jetty where we secured her. Whilst attempting to dismantle her on the water, she heeled over, filled up and sank. We eventually hauled her out of the water in pieces, which were put on a trailer and taken back to Grain by road.'
   Interest in the Cork was shown in many quarters, particularly by the American Aviation Mission in August 1918, and by Boulton & Paul Ltd who received drawings of the hull in December 1919 to study for its conversion to steel and wood construction.
   A civil variant of the Cork, designated P.8, was considered by the Phoenix Dynamo Co. It was to carry ten passengers or freight. Manning also investigated the possibility of entering the Cork for the ?10,000 prize offered by the Daily Mail for the first direct transatlantic flight. His calculations indicate that the Cork could have accomplished the flight with ease.
   The Cork was later developed as the Napier Lion-powered English Electric P.5 Kingston.

   Span: upper 85 ft 6 in, lower 63 ft 6 in; length: Mk I 48 ft 7 3/16 in, Mk II and III 49 ft 1 11/16 in; height: Mk I 20 ft, Mk II and III 21 ft 2 in; length of hull 45 ft; hull width across chines 7 ft 6 in; hull maximum depth 6 ft 2 in; wing chord 9 ft; wing incidence 40; wing dihedral: upper 3 (outboard section only), lower 0; gap at inboard interplane-struts 10 ft; stagger nil; aileron span 19 ft 11 1/2 in; aileron chord 2 ft 1 in; tailplane span 25 ft; tailplane chord including elevator 8 ft 6 1/2 in; tailplane normal incidence 5 30'; elevator span 25 ft; elevator chord 2 ft 9 m; fin height 6 ft 3 m; fin root chord 10ft 1 in; rudder height Mk II and III 10 ft 3 5/8 in; rudder chord Mk II and III 4 ft 0 1/8 in; propeller diameter 10 ft; wing area including ailerons: Mk I 1,292 sq ft, Mk II and III 1,340,5 sq ft; total aileron area 85,5 sq ft; tailplane area including elevator 200 sq ft; total elevator area 58 sq ft; fin area 31 sq ft; rudder area: Mk I (9 Aug, 1918) 22.5 sq ft, Mk II and III 42 sq ft; total stabiliser area 34.5 sq ft.
   Hull weight: P.5 1,417 lb, P.5A 1,306 lb; empty weight: Mk I 7,350 lb; loaded weight: Mk I 11,600 lb.
   Maximum speed: Mk I 105 mph at sea level; climb to 5,000 ft: Mk I 10 min; service ceiling: Mk I 13,000 ft; endurance: Mk I 8 hr.