This aeroplane, which looked almost like a fossil bird, was a result of the request for a two-seat anti-submarine patrol aircraft from the Greek Admiralty, and was originally to have been a floatplane. It was one of the first of T. Sopwith's aircraft to be designed specifically for military use: an order for three with dual controls was placed by the Greek government in 1913, with a second order for six being placed in February 1914. These were delivered as land planes with a four-wheel undercarriage between July and October 1914, but most were commandeered by the British Admiralty for use by the Royal Navy Air Service (RNAS). These were numbered 801 - 806 and were named the Admiralty Type 806. In fact the British Admiralty had ordered two machines in October 1913 and these were delivered to the Isle of Grain in March 1914.

        The term "gunbus" is frequently associated with the Vickers FB5 and FB9 fighting biplanes, but it was a generic term used at the time to describe any armed scout or reconnaissance aircraft. The pusher design was adopted because there was a lack of a reliable interrupter gear which would stop the pilot from shooting off his own propeller. Instead the observer/gunner had a wide field of fire and in the case of the Sopwith Gunbus a 0.303 inch Lewis machine gun was sometimes mounted on a telescopic pillar.

      In June 1915 another 30 machines were ordered by the British Admiralty from Robey and Co of Lincoln, but these were slightly different from the earlier examples, with a new nacelle, a two-wheel undercarriage with a narrower track, and cut-outs for the boom attachment points on the wing trailing edges. Some of the Robey machines had a transparent hatch beneath the front cockpit and racks for four or 6 x 65lb bombs under the lower wings just outboard of the second pair of interplane struts: it is unlikely that any of them were used as operational bombers. The pilot was moved into the front cockpit in the Robey machines, but only 17 aircraft were delivered complete with the remainder being delivered as spares. These machines were powered by a 150hp Sunbeam Napier 8 cylinder engine - a water cooled V more usually used in airships.

       How many of these machines saw operational service in France is unclear: at least one was used for a short time at Dunkirk by Commander Samson's RNAS squadron where keeping it airworthy seems to have caused considerable headaches. This, combined with a top speed of 80 mph, is probably why the remainder were used for training at Hendon, Detling and Eastchurch until the beginning of 1916, when they had been written off either by accidents or by damage caused by stormy weather.  

      Like many other little-known early aircraft types there is no model of this aircraft produced in 1/72 scale, although card models in 1/48 and 1/50 scales are available. The only reference to a conversion that I know of was by G. Scarborough published in Airfix Magazine Annual No 5 in 1975 and I used this as the basis for my model. G. Scarborough did not build a full aircraft as his was part of a small diorama with the aeroplane partly dismantled and being towed on a trailer, as illustrated in a series of photographs of Robey built machines being taken from the factory in Lincoln to a nearby testing field. In this diorama the wings and horizontal tail unit were carried in the towing lorry. I have chosen to represent the entire aircraft without any vehicles. There are two (possibly more) kits that could be used as starting points: the Airfix Avro 504 (as in my model), or the Airfix DH 4, both of which could provide wings, wheels, struts, etc, but in either case you will need two kits unless you are prepared to make part of the wings from card. I used G. Scarborough's method of increasing the chord of the Avro 504 wings and kept the fuselages for later use to correct the errors in that kit when I make different variants of that famous type. The engines and propellors also proved useful for the conversions of Vickers Gunbuses from D H 4's: it is surprising how much of a kit can be used in different conversions.

    Start by making the fuselage nacelle. This was built from 20 thou card sides and floor with three bulkheads, one at the point where the top decking starts to curve at the nose end, one between the cockpits and one at the rear. Construct a three sided box with the sides, floor and bulkheads and add any cockpit detail including seats, and a control column, rudder bar, etc in the front cockpit. The top decking was from 20 thou card bent over the curved tops of the bulkheads and glued into place with liquid cement. Cut a slot 20 thou wide in the top decking, 2mm behind the rear cockpit to take the radiator which was made from card 11mm x 12mm with the surfaces scored to represent the radiator grill. The nose of the nacelle was built up with scrap plastic and then covered with filler and sanded to shape. The coamings around the cockpits were made from thin rod or stretched sprue. The engine bearers were 15mm lengths of rod inserted into holes drilled into the rear of the nacelle. The rear support for the engine was cut from 20 thou card and cemented to the rear ends of the engine bearer rods. Thin rod or sprue made the additional arms of the bearers, and card was used for the brackets at the front end. The nacelle and engine bearers were painted before the engine was cemented into place. Drill two holes under the nacelle to take the undercarriage legs.

    The engine was difficult to make as I could not find any good quality photos or drawings: the drawings provided in the AM Annual were poor and not very accurate but were among the better ones that I could find, so they provided the basis for my efforts. Make the cylinder blocks from two pieces of 60 thou laminated card with the ends rounded off. The engine block/sump was also made from two pieces of 60 thou card, with a 20 thou piece sandwiched between them. When this was dry I sanded to shape the bottom and rear end, making sure that it fitted snugly into the space between the engine bearers. The top of the engine block should slope inwards to take the cylinder blocks. Drill a hole in the rear of the engine block for the propellor. Assemble the cylinders to the engine block and then add the exhaust pipes from short pieces of rod bent to a curve. I put some very thin rod on the sides of the block to represent pipes, and other tiny details as my imagination allowed. The assembly was painted light grey with a little silver mixed in to give a metallic finish. When the engine is finished cement the radiator to the nacelle and the engine to the bearers, and then add the water pipes from the cylinders to the radiator. I painted the water pipes copper with some brown mixed in to take off the sheen. The exhaust pipes were rust. The propellor blades were wide and had to be made from card and glued to the boss of an old Camel(?) propellor which I found in the bottom of my spares box which had had the original blades removed. (Why I had kept this item I cannot remember: such are my hoarding tendencies). The propellor was painted dark wood with a grey boss.

      For the wings you will need sets from two kits although I only used three wing units. I used Airfix Avro 504 wings but the chord of the Sopwith Gunbus was greater than that of the Avro so this had to be increased as explained later. Start by cutting a length 8.4 cm from the port (left) side of both upper wings: this will extend past the centre section cut-out almost to the point where the wing anhedral starts on the starboard (right) side. Score the underside of both wings and remove the anhedral by carefully bending the wing until it is flat: secure the new shape by running liquid cement along the score lines and press flat until the wings are dry. Now take one of the bottom wings and starting from the port (left) side cut off a piece 4.9 cm long. Cut a similar length from the starboard(right) side of the lower wing. These will form the centre sections of the new wings. Finally cut two lengths 1.1 cm long from what is left of the inner sections of the lower wing.

    To assemble the new wings carefully cement the two 1.1cm pieces to the starboard (right) sections: these should be butted at 90 degrees and allowed to dry thoroughly. When they are completely dry file the edges of the centre section pieces (the ones 4.9 cm long) and the edges of the outer wing panels so that when they are butted together the correct dihedral angle (3.5 degrees) is achieved. This part is a little tricky so work carefully as the look of the model depends on this being as accurate as possible. The top wing should now have the cut-out for the booms removed. Do this by cutting the slots 6mm deep x 5 mm wide where the anhedral starts: these should be removed from the centre section. When you have done this join the wing sections ensuring that you get the correct dihedral angle on both sides and to do this support the wing while it dries out. With the bottom wing the centre section chord needs to be 1.6 cm, so reduce the chord to this dimension before you glue the wing sections together by removing a piece from the trailing edge: do not file down the trailing edge as this had a slightly thicker section. Now assemble the lower wing sections but in this case allow the wing centre section leading edge to extend 2mm ahead of the outer wing sections. This will allow you to put on the rod on the outer wing sections only when you later extend the chord of the wings. Once again check that you get the correct dihedral and support the wing while it dries out.

    When the new wings are completely dry, glue a length of 60 thou diameter rod along the entire leading edge of the upper wing and the outer panels of the lower wing and allow to dry. Fill the gap between the rod and the top and bottom surfaces of the wings with filler, and at the same time fill any strut location holes, the strut grooves in the underside of the top wing with card and filler, the old aileron grooves and the joints of the wing sections. I also filled the cutouts of the 504 wing centre sections with card and filled the joints. When all the filler is being rubbed down you can also remove the ribs of the 504, and shape the tips of the wings. New ribs can be added to the upper surfaces of both wings by scoring shallow grooves and putting in very thin rod or stretched sprue with liquid cement. When this is dry lightly sand the new ribs until they are nearly flush with the wing surface, and then score new ailerons on both wings. Mark and carefully drill the new strut location holes on both wings, and on the lower wing drill holes for the undercarriage under the centre section, and wing skids under the outer struts. Last of all you will need to cut and file shallow grooves on the top of the trailing edges of the wings where the booms will be joined later. Strictly this is not accurate as the booms were inserted into the rear of the wings but this is not really practical on this model as the wings are so thin that the resulting joint would be very weak.

     I made my booms from wire as I have found that these are very important structural elements in pusher models and at this scale strength really matters. The booms are 8.7 cm long and the end which will sit on the top of the wing needs to be rounded off with a file. The rear ends need to have chamfered faces so that they will butt together properly to form a neat V. Glue the booms to the tops of the wings (I used epoxy glue for this operation), and the V joints with superglue. Support the wings and booms while the glue dries: the angle of the lower booms relative to the wing trailing edge is quite large and needs to be right or the model will not sit properly. For this you will need to consult the plans. The angle of the booms on the upper wing is much less but again care is needed to get this right, but it is worth the time and effort required and is not difficult to achieve. When the boom joints are dry cut the cross struts from card shaped to airofoil section and glue into place with superglue. A small piece of thin rod represents the third strut which supports the base of the fin on the lower boom. Cut a fin and rudder from 20 thou card sanded to shape and airofoil section. The horizontal tail surfaces were also cut from 20 thou card. A small section was cut out from the top rear of the fin so that the boom wires would fit. Score the fin and elevators with a sharp knife. Drill a tiny hole in the base of the fin 2-3 mm forward of the rudder line and insert a pin from thin rod 5 mm long. This will fit into the V of the lower boom and help to hold the fin in place during the final assembly, and hold the tail skid.

      I painted the model at this stage as otherwise it would not be practical to do so. The upper surfaces of the wings and tail were PC8 (three parts Humbrol 81 and two parts Humbrol 26) while the under surfaces and sides and bottom of the nacelle were clear doped linen. The top of the nacelle, the fin and the leading edge of the rudder were light grey, with a white band on the fin where the serial was located. The tyres were dark grey but the wheel discs on some machines were white when they left the Robey factory. These machines also had the serial on a horizontal white band on the fin. The struts and booms were Revell semi-matt 382 but were not painted until after the final assembly was completed, and the radiator black. Markings on these machines varied greatly as photographs show, so select the machine you wish to model and use the photos to guide you. I based mine on 3842 as it probably appeared in late 1915: photos of at least one, possibly two machines show them with RNAS roundels under the wing tips and rudder stripes and I assume that 3842 had similar markings. I painted RNAS roundels on the upper wings: the photos clearly show the white outer rings. To hand paint roundels scribe the outer circle with a pair of dividers and then paint the inside of the circle white. When this is dry scribe two concentric circles, again with the dividers, for the red ring and then paint this. The paint will run into the grooves made by the dividers and give you a nice clean edge, but you will need a fine brush, a steady hand and thin paint. You may need to use two coats of paint especially for the white. Alternatively you could use RNAS roundels from the Pegasus transfer sheet early RNAS roundels and flags (PGS 72 001) which I managed to find only after I had completed my model. I also painted the white stripe on the fin for the serial but whether this survived in service I cannot be sure. The serial came from dry rub-down transfers and the rudder stripes were hand painted.

     Fitting the nacelle to the lower wing and then the upper wing to the nacelle was a lot less difficult than I had feared it would be. If you take care and are patient it can be achieved without tears. Some people may want to use a jig but I did not find this necessary. You will need to work fairly quickly but you must also be careful so plan everything beforehand and have all the parts and glues ready before you begin. Allow about half an hour for this operation. Start by measuring the exact lengths of the nacelle struts and then cement them to the sides of the nacelle. The top of the radiator acts as a guide as to where the tops of the struts should be - allow for the struts to go into the holes in the wings. When you cement the struts to the rear of the nacelle make sure that they are at the correct angle for the stagger of the wings. When the struts are drying insert the fin/rudder into the rear of the booms of the upper wing assembly and glue into place with superglue, making sure that the fin is vertical and aligns correctly fore and aft. Now put a drop of glue on to the bottom of the nacelle struts and carefully lower them into the holes in the lower wing. Put more drops of glue on to the tops of the nacelle struts and run a little along the top of the radiator and then gently lower the top wing so that the leading edge of the top wing aligns with the edge of the radiator and the nacelle struts enter the holes on the underside of the wing. The pin on the bottom of the fin should drop into the apex of the V of the lower boom: it can be glued with superglue. Although this structure sounds weak it is stronger than you might think! Carefully cement two struts into the outer holes of both sides of the wings: these came from one of the 504's and should be the correct size without needing adjustment. This sub-assembly can now be allowed to dry thoroughly and you can go off and have a celebratory drink and relax as the most difficult part is now over. Although this assembly will not stand rough handling, when dry it can be manipulated sufficiently for the remaining wing and the boom struts to be inserted with few problems. When it is dry the remaining four struts from the second 504 kit can be inserted into the wings and the rest either raided from the spares box or made and shaped from card. The boom struts were made from card and inserted using superglue to hold them in place. I inserted at least two sets of boom struts before finishing the wings simply to make the structure a little stronger. When the assembly is complete it is much stronger than you think, but I do not recommend any test flights!

    The undercarriage is next. Cut the struts from card or Evergreen strip and shape to airofoil section, and cut the skids from the same material and shape the tips. Drill small holes in the top of the skids where the struts will fit. I cut an axle from wire but thick plastic rod would also do. Cement the axle to the skids and allow to dry. Cement the struts into the tops of the skids making sure that they are angled inwards with the front ones vertical in relation to the skid and the rear ones angled forwards. While the glue on the struts is still soft put some glue into the holes under the nacelle and wing and carefully lower the undercarriage unit so that the ends of the struts are inserted into the correct holes. Adjust the struts and skids so that the struts are at the correct angles and the skids are parallel with the bottom of the fuselage and set aside to dry thoroughly. Glue the wheels on to the axle with superglue if the axle is wire, after the undercarriage sub-assembly has been painted. Cut control horns from thin card and glue on to the flight control surfaces and below the pilot's cockpit on the side of the nacelle. All of the struts, booms and horns can now be painted.

     Now for the fun part; rigging. There are about 140 individual wires to be put on to this model (more than any other that I have made, and compared with just over 50 on the DH 4), so expect it to take some time to complete. Start by rigging the front to rear wires between the wing struts, working from the centre outwards. Then rig the rear of the wings followed by the front. The sides of the booms are next and when the rear bays of the booms have been rigged the horizontal tail surfaces can be glued into place. Rig the horizontal wires between the booms but before you complete the forward bays put on the propellor. The undercarriage is next and then the control wires on the wings and tail, the control wires from the horns on the nacelle to the struts, and the tail bracing. Finally add the elevator and rudder wires, the anti drag wires, and the balance wire along the leading edge of the top wing. The tail skid and wing skids (the latter made from 5 amp fuse wire) complete the model.

      This is very much an aircraft of its time - it looks ungainly and fragile yet as a model it is quite strong and in real life it seems to have served reasonably well in the unglamorous role of trainer. Like many other early aircraft it lacked refinement and so it adds an interesting touch to any collection. It is not surprising that it has not attracted a great deal of attention: pusher aircraft generally are not popular with manufacturers, nor it seems with many modellers either. I wanted to build one of these when I first saw it in the AM annual but I never got around to it, and then gave up modelling for over 30 years. Of all the pusher conversions that I have built so far I expected this to be the most difficult because the wings are not directly attached to the fuselage, the large number of struts and the stagger on the wings all of which present particular problems. However I was surprised at how few difficulties I actually encountered when I came to make it. I would certainly not recommend it to anyone who has not previously attempted biplane conversions or had experience with pusher designs, but it is an interesting challenge and was not as difficult as the description might make it seem. I am not an OOB modeller, I prefer scratch building or conversions as they offer more of a challenge: if you are like me this could be for you. You could also have a unique model in your collection.

Stephen Foster