The v-twin for sprint work

UP to last year the V-twin engined racing-car was still very much in evidence at speed hill climbs and speed trials. Usually referred to as a ” Shelsley Special”—wrongly we think, because such cars rim at other important venues apart from the Worcestershire hill and because four and even sixcylinder engines are installed in chassis which equally deserve this designation— these air-cooled, two-cylinder racing-cars have, of recent times, been developed to compete seriously against contemporary conventional racing-cars. Their advantage now lies in low weight, due to the absence of water cooling mediums and economy of components, and a good output by reason of attention to carburation and combustion details allied to the hereditary efficiency of the motorcycle type, both mechanically and volumetrically. In the early days the V-twin and invariably chain-driven racing-car owed nothing to specialised devotion and development. The G.N. of post 1914-18 days adopted chain in place of belt transmission and turned its engine round to a transverse position across the frame, and became a sporting proposition in ” Vitesse ” guise. Racing was the thing to do amongst manufacturers in those days, and so Godfrey and , Nash built racing versions. As early as 1920, the single-seater G.N. ” Kim” was reaching about 90 m.p.h. on Brooklands and the early J.C.C. 200 Mile Races led to the construction of special cars having o.h.c. engines, of the ” 3ilowrdi ” and ” Akela ” types. Capt. Frazer-Nash, to whom we owe so much in the way of German bomber destruction in this war, won the 1,100 c.c. class of the 1921 ” 200 ” in his G.N., at 71.54 m.p.h. In 1922 a team of three o.h.c. G.N.s ran, but they were vanquished by the four-cylinder Salmsons. The G.N. was also prominent in sprint events. B. H. Davenport upheld the reputation of the V-twin racing cyclecar after the conventional racing-car had ousted it from most other spheres, by taking the record at Shelsley Walsh in 1926 in 48.8 secs., in 1927 in 47.8 secs., and in 1928 in 46.2 secs. These exploits of Davenport with a car usually described as a Frazer-Nash, but in reality built up of G.N. and special components, opened up the era of renewed interest in the development of the V-twin engine and G.N. type of chassis for sprint events. In the early post-war days, the twincylinder engine was just as logical in a small racing-car as a multi-cylinder unit. Then, as the four-cylinder engine became universal for ordinary work, the air-cooled, racing cyclecar faded out. Its revival commenced more or less mildly, as youthful enthusiasts, fascinated with the type and able to purchase touring G.N.s very cheaply and strip them, raced such cars as Edwardian enthusiasts race pre-1914 cars, for the fun of it, rath:r than in direct competition with later types. Over it all, the achievements of Davenport must have shone out, for from about 1930 or so onwards, a marked revival of interest in the V-twin racing-car has been evident, and, in spite of dope fuels, supercharging, and such like, it has been subjected to specialised development for direct competition with 1,100 c.c. and 1+-litre cars of all kinds. That its certain inherent fandamental advantages can be employed to good effect must be evident to its followers from the ability of Davenport’s car to beat contemporary Grand Prix Sunbeams, Vauxhalls, Brescia Bugattis, and Beardmores, etc., during the period 1926-8. Before going on to survey some of the recent V-twin racing jobs, it is interesting to take the 1920 racing G.N. as a standard by which to gauge subsequent development work on the air-cooled V-twin engine. The G.N. used at this time by Capt. Frazer-Nash for track racing had an engine of 84 mm. x 98 mm. (1,086 c.c.) designed and built for the 1914 Cydecar G.P., and modified somewhat in 1919. It was a 90° twin with vertical finning on the heads, which were each held, down by four long studs. The valves, 47 nun, in dia. and of ordinary nickel steel, were set at 90° and operated by push-rods and rockers. There was a curved bridge member over each head, carrying an auxiliary valve guide, the valve spring abutments, andh.t. lead tube. The plug entered the head at an angle between the valves, with the compression tap opposite. Ignition was by a single chain-driven magneto. Lubrication was by twin lines from the tank, one feeding the main bearings, the other the crankcase interior. The flywheel was steel, and the big-ends bronze. A single carburetter was used, and the cams were special. With knife-edge nose and long streamlined tai!, this engine pulled ratios of 3, 4+ and 7 to 1, with 650 nun. X 60 mm. tyres. Incidentally, the chassis had Wire and bobbin steering, three leaf quarter-elliptic front springs with radius rods below, and two leaf damper springs above, and quarter-elliptic rear springs with Houdaille shock-absorbers and additional damping from ash-buffers carrying rubber pads. Davenport’s car, known as “The Spider,” had an 89 mm.x120 mm. (1,493 c.c.) 90° ergine, with four valve heads, and chain-driven o.h. camshafts. The pistons were Hepolite, Ore carburetter a Solex, and ignition was by M.L. magneto. The car ran on Discol R.D. fuel and Castrol oil. It cost its owner £300 in pieces, and something like the same sum was spent on development. Apart from its three absolute records at Shelsley Walsh, it was raced at South Harting and Holme Moss in 1924, and at Southport, Blackpool, Colwyn Bay, Morecambe, Skegness, Gopsall Park and Stalybridge, and at Audenshaw dirt track. It made fastest time of the day on more than thirty occasions, and came within 1½ secs. of the solo motor-cycle lap record when tried on the cinders. “Spider” used to cross the line at Blackpool at 100 m.p.h., and reached 97 m.p.h. at Southport, while some idea of its get-away is obtained when it is mentioned that it has been started in top speed at Southport and still beaten all rivals. The car was bought from Capt. Frazer-Nash in 1922, and built up in 1923. Davenport subsequently designed many new parts, and rebuilt the engine about a dozen times, until only the heads, timing case cover and one cam-box remained of the original. Victor Horsman, the Triumph exponent, designed special big-ends. The chassis was shortened one foot, and a single-seater streamlined body made up. As the old Southport rules demanded a passenger, a two-seater body was also fitted on occasion, as carrying a passenger on the tail was found to be too precarious! The weight was approximately 7½ cwt. For Shelsley record runs the ratios were 5.2, 6.4 and 8.9 to 1, but for Southport and Blackpool, four speeds were used, of 3.5, 4.8, 6.4 and 8.9 to 1. The engine gave excellent power over a wide speed range, the maximum being about 75 b.h.p. at 5,000 r.p.m. A maximum of some 65 m.p.h. was available on the 6.4 to 1 ratio. The cylinders were off-set.

Of recent years, H. G. Symonds, who used to have offices in the same building as the old MOTOR SPORT offices, revived “The Spider,” and ran it in sprint events, but it was not running as well as when Davenport had it.

Another racing G.N. developed from the original 200 Mile Race cars, and very prominent at Southport was E. L. F. Mucklow’s “Tiger III.” This had the “AkeIa” 84 x 98 mm. (1,087 c.c.) V-twin G.N. engine, with the shaft-driven o.h. camshafts. Spindles driven from the crankshaft carried the base bevels for the drive, and the vertical shafts in front of the cylinders had vernier couplings to facilitate valve timings, and were splined at the top end for head removal, while the heads were attached to the crankcase by four long studs apiece, and the top camshaft bevels enclosed in aluminium cam-boxes—normal G.N. “Akela” practice. There was a large Solex carburetter per cylinder, fed by air-pressure from a tank in the tail. The compression-ratio was approximately 9 to 1, and the output was in the region of 55 b.h.p. Ignition was by two Bosch magnetos driven one from each timing spindle. Lubrication was by hand-pump to the top and bottom timing gears by separate feeds, and from the same tank by pressure to the crankcase. Four speeds were used, and the single-seater body resembled those of “Kim” and “Spider,” with taper nose, cylinders protruding, and detachable undershields ran from front to rear axles. Apart from variation of sprockets, further change of gear-ratio was possible by using two different rear wheel sizes. This car was consistently raced by H. B. Prestwich, but suffered minor engine damage when a piston broke, around 1933.

E. J. Moor went into the matter of developing the G.N. engines very thoroughly indeed, with his series of “Wasps,” one of which he was still racing last year. “Wasp I” was built in 1924, just when the more conventional racing-cars were compelling the normal racing cyclecars to take a back seat. The engine was the 84 x 98 mm, chain-drive o.h. camshaft twin, having two valves per cylinder. Here we may digress to say that we have always considered this unit to be the “Mowgli.” However, the issue is confused because some of these chain-driven o.h.c. engines have two valves per cylinder and others four, although in this case “Akela” heads have probably been adopted to take sprockets on the camshafts instead of bevels; even H. R. Godfrey has forgotten the type names of the various G.N. engines, and we should like to hear from anyone who knows them. These chain-drive o.h.c. engines have the drive behind the cylinders, the long ordinary-link chain driving a camshaft over each head, which operates the valves via rockers enclosed in alloy cam-boxes, while the chain is driven direct from an external sprocket on the crankshaft and tensioned by a jockey-sprocket which slides for adjustment in a slotted bracket on the top face of the crankcase, and is locked by a clamp bolt. “Wasp” had one of these engines considerably tuned, and a crude single-seater, yellow and black body replacing the original two-seater. She ran well in North Country events, although tending to inflammability. Later new cylinders and a longer stroke of 89 mm. x120 mm. (1,493 c.c.) was tried. The output was improved, but the cylinder holding-down studs now tore out of the alloy crankcase. A new cast phosphor-bronze crankcase, to some extent lightened by turning, was made in London, and the engine then commenced to break connecting rods. It is most interesting that the crankcase never suffered, although the timing gear, valves, and rockers would be badly damaged as the pistons contacted the heads. A cure was effected by making up new con-rods of Vickers-Armstrong “Vibrax,” heat-treated to 110 tons tensile. At this period, Davenport’s four-valve “Spider” was faster than the “Wasp,” and also often fastest amongst the 2-litre cars. Consequently, Moor increased the bore by ½ mm., to 89½ x120 mm., or 1,510 c.c. This enabled him to enter the 2-litre Class, and in 1928 G.N.s were first in classes 2, 3 and 4 at Shelsley. Thereafter the supercharged 2-litre Bugattis really began to move, but “Wasp” still managed second place in the class. “Wasp I” was raced until she came off her tow-bar and split a telegraph pole right to the top; she was eventually rebuilt, but always suffered from a bent chassis. “Wasp II” came into being in 1930. She was one of the “Akela” 200 Mile Race cars, in poor condition. Camshafts and rocker-arms from Davenport’s 1929 engine were installed, a new big-end fitted, and the body lightened. In this form she was about 4 secs. faster at Shelslev than “Wasp I.” During the winter, thorough rebuilding took place. Two track-racing Amal carburetters, a different type of big-end, and Rudge motor-cycle dirt-track pistons were used, and a very light, tailless body fitted, the driving chains being quite exposed. The wheelbase was reduced by one foot, and a new front axle, 9″ wider, fitted. On her first appearance “Wasp III” climbed Shelsley in 48 secs., a 1,100 c.c. class record.

At the next Shelsley meeting, three valve springs replaced the original two per valve. It was on this occasion that the off-side cam-box blew off at the S-bend, rockers, tappets, bits of valve, and so on, sailing into the trees. The climb was completed in 57 secs., and all the pieces subsequently retrieved and welded up in time for the car to run in the Brighton speed trials. On this occasion faulty dope resulted in momentary flat-spots, but the standing half-mile was covered at an average of 56 m.p.h. That winter a tail was added to the body. At the first Shelsley meeting of 1933 the engine mysteriously lost 1,000 r.p.m., eventually traced to a broken inlet rocker arm, possibly dating from the previous year’s disintegration.

“Wasp IV” had a Brescia Bugatti engine, and is outside the scope of the present study. “Wasp III” was rebuilt for the 1934 season. The original “Akela” bronze four-valve heads had been very materially modified and had reached the limit of efficiency. So Norton racing 596 c.c. heads were adapted to the G.N. engine. To accommodate the difference in holding-down bolt centres, the original holes were plugged, the crankcase re-drilled and tapped, and the barrels freshly drilled. As there was an appreciable difference in the camshaft drive lines, new shafts were made up, roughly 1″ longer, and the drive end of the Norton cam-boxes was turned off, and aluminium castings made, welded on to Norton boxes, and machined to take the G.N. drive. Adaptors were made to enable the Amal carburetters to be used with the 60º flanges. One set of cams had to be reversed and the plug hole tapped in the opposite side of the head, to further suit the Norton heads to a 90° twin-cylinder engine. The engine now easily ran up to 7,000 r.p.m. It was raced until quite recently, but in 1,066 c.c. form.

One of the first attempts to supercharge a V-twin car engine, and one of the few attempts to blow a G.N. unit, was made by H. B. Showell about 1932. In 1927 Showell bought one of the 1922 200 Mile Race G.N.s which had been enlarged to 1½-litres, with single carburetter and magneto. In 1928, with two carburetters and raised compression-ratio, it ran at Shelsley and crashed badly, damaging crankcase, front axle and chassis. Some £300-£400 was then spent in rebuilding it. The 1,493 c.c. engine had phosphor-bronze four-valve heads and the chain camshaft drive before the cylinders— yet another version of G.N. layout. The compression-ratio was 9.8 to 1 unblown, and the engine peaked at 5,000 r.p.m. Later a No. 8 Cozette compressor was driven from the nose of the crankshaft. It drew from an S.U. carburetter bolted direct to the casing at frame level on the off side and fed into a T-pipe attached by hose joint to the outlet on the near side, the blow off valve being beneath this joint. This T-pipe joined the ribbed inlet ports by hose-joints. Two magnetos were used with two plugs per cylinder, hung from plates bolted to each rocker-box, and driven from the camshafts, the rear plug leads passing through a guide clipped to the induction pipe. The advance and retard levers were coupled by a transverse rod, spring-loaded to the off-side mag. plate, and controlled by Bowden cable. The car never appeared prominently. About the same time, Keith Anderson built the blown G.N. “Akela II,” but this, too, never made a name for itself. That cheery mortal, Edward Pyddocke, also built up a “Mowgli” G.N. and ran it on the road in full sprint trim for a while.

Breaking away from usual G.N. practice, but achieving greater success than almost any other racing V-twin of the period, John Bolster’s Bolster-Special ″Bloody Mary,” did much to remind sceptics that reduction of weight by reason of air-cooling and a simple transmission, could embrace performance unbeatable by modern racing jobs. The Bolster-Special was evolved from G.N. and Bleriot-Whippet bits, and was run as a road car for a brief, but exciting, spell. She was built in 1929, with a pre-war J.A.P. engine, an orphan gearbox and belt drive. When the first engine had had enough and had complained in no uncertain terms, another was put in, which hauled out the crude transmission. A 1924 s.v. four-cam 981 c.c. J.A.P. V-twin was put in, set in line with the wooden frame of 1½” x3″ ash, driving by chain to a four-speed close-ratio Sturmey-Archer gearbox, with final chain drive. “Mary” climbed Shelsley in 48 4/5 secs. in 1932, and a 1926 o.h.v. J.A.P. motor was then put in. This was quite a standard push-rod 981 c.c. engine, with polished internals, raised compression-ratio and special balance, the maximum revs. being about 6,500 r.p.m. The ash chassis was carried quite high on quarter-elliptic springs, and the centre of gravity kept reasonable by seating John alongside the machinery on a low-slung plank, the solid rear axle just behind him. There was a single brake on the rear axle, and the driver used to bind his left arm as some protection from the close proximity of exhaust and chains. The gear-ratios were changed to suit different courses, by fitting different engine sprockets and gearbox pinions. For Shelsley the gear-ratios were: 6.1, 7.3, 9.1 and 11.8 to 1, top being used up to the “S” and again on the final straight, and second right through the “S.” The whole car weighed only 5 cwt. It performed well and sensationally at many venues, although overturning at Shelsley in 1938. For 1934 Bolster wanted better road-holding, and f.w.b., which meant more avoirdupois. To counter this, more power was needed, but to further increase the engine speed, or to supercharge, was considered likely to court unreliability. The problem was very cleverly solved by lengthening and strengthening the chassis and adding another J.A.P. V-twin, giving a total capacity of 1,162 c.c. The engines were mounted in tandem, and coupled by a jockey-tensioned chain. The rear engine drove by chain to the gearbox, and the gearbox by chain to the rear axle. The car ran very well indeed, and was only laid up when Bolster borrowed its two engines for his new four-engined, 4-litre job, in which a metal chassis, more conventional driving position and i.f.s. were incorporated.

This car, although using V-twin engines, comes rather outside the scope of this thesis, as do the various specials using a multitude of single-cylinder engines, such as Charles Martin’s ambitious car with four J.A.P. engines set in a square, and. Richard Bolster’s fairly recent experiment with four motor-cycle engines set in line in his G.N. chassis, and Waddy’s F.W.D. two-engined Fuzzi. Such cars call for special means of coupling the engines and special chassis, and are not only costly, but highly experimental, so that they are just as likely to be temperamental in the Paddock as Bolster feared his J.A.P. engine would have been had he blown it—admitted his solution was to multiply motors, but it takes a Bolster to do such things successfully. So the author will confine himself to developments of early V-twins, or to chassis in which well-proven proprietary V-twins have been easily installed.

Richard Bolster’s G.N. ran in 1933 with a “Mowgli” 1,086 c.c. engine having twin carburetters, special heads, polished ports, h.c. pistons and roller big-ends. It ran on dope fuel and would do ten practice laps at Donington and a couple of races without trouble. The chassis was a foot shorter than standard, had four-speed G.N. transmission, Hartfords fore and aft and a special f.w.b. front axle. Later, a six-cylinder M.G. Magnette engine went in, cooled by a Darracq radiator, but we are only concerned with two-lungers.

Another racing V-twin which reached its prime around the 1932-4 period is the Meo-Special, interesting because it was prepared for Brooklands racing rather than for speed trial work. M. Meo commenced conversion work on a 1922 G.N. with “Mowgli” engine in 1929. The two-valve 84 x98 mm. engine was considerably modified by 1932. It had I-section connecting rods, and the main and big-end bearings used ⅜” x ⅜″ uncaged rollers. The heads were Meo’s own, having almost hemispherical combustion chambers, with 14 mm. plugs set towards the centre of the head and inlet valves larger than the exhausts. A large Amal carburetter per cylinder replaced the original, and these were mounted with chokes at 60º to horizontal to give a straight flow into the ports. The “Mowgli” chain camshaft drive was replaced by separate chain drive to each camshaft, the chains being enclosed and tensioned by leaf springs. The rockers actuated the valves via short push-rods to relieve side thrust. The compression-ratio was 9 to 1, but could be raised to 10.5 to 1 by removal of compression plates. The special pistons were of “Y” alloy and motor-cycle hairpin valve springs were used. The engine ran on R.D.1. Ignition was by a Bosch magneto driven by vernier coupling from the nose of the crankshaft, and a cut-out was carried on the gear-lever. The engine was very considerably modified, even to the method of cylinder barrel and head anchorage, and finally only a few 1922 components remained. The chassis was likewise materially altered, until only the bevel-box and a few odd pieces of the original were in use. The new frame was rigid, unlike the usual G.N. “bedstead.” From just behind the clutch to a point in front of the bevel box, two channel members formed a large “A,” and at the apex of the “A” were two cross-members, one below and one above the chassis side-members. Straight side-members were retained. Suspension was by quarter-elliptic springs mounted on steel blocks about 3″ above the chassis to give a very low build. The drive passed via a G.N. type clutch and bevel-box to four-speed G.N. transmission, giving ratios of 3.5, 4.0, 5.8 and 8.3 to 1. The outside gear-lever had a locking device to prevent the engagement of two speeds at once, and the chains were 3/4″ x 7/16″. The wire wheels carried 700 x 80 tyres. The tubular front axle was of straight type, which gave a large span between spring anchorages. The shock-absorber arms acted as radius arms, which would be relieved of much of the braking torque had f.w.b. been fitted. It was planned to fit such brakes, operated by steel strip. Steering was by special G.N. bevel-box and drop arm. The front track was 3′ 9″; that at the rear 3′ 4″. A very well streamlined wooden two-seater body was fitted, with staggered hammock seats, cylinders within a bonnet from which only the cam-boxes protruded, and a long tail and full undershield. The exhaust pipes ran down each side of the body, the ports having finned coolers, into Brooklands silencers. The facia was cowled before the driver, and carried rev. counter and air-pressure gauge, and the steering wheel was spring-spoked. The 6½ gallon fuel tank in the tail was fed by hand-pressure and had a Meo quick-action cap. The engine was lubricated by a Pilgrim pump. The Meo was entered for the 1932 B.A.R.C. Inter-Club Meeting, but was a non-starter due to slip-ring trouble. The writer had a brief ride in it during practice, when it did about 90 m.p.h. over the quarter mile. It ran again at the 1933 Inter-Club Meeting, but was over-handicapped. Since this time it has disappeared, and it would be most interesting to know where it now is, for it was about the last of the track-racing cyclecars.

Another remarkable job was the Glegg brothers’ “Dorcas,” which had a J.A.P. engine and four-wheel drive, composed of a B.S.A. front assembly and Bugatti rear parts, the drive being arranged to change from four to two-wheel drive while cornering, at one time. “Dorcas” once assayed an M.C.C. “Exeter” and it still runs at sprint meetings, now very well bodied.

Typical of modern cyclecar practice is David Fry’s remarkable “Freikaiserwagen.” Largely designed by Dick Caesar, this car was built at Cambridge in 1937. The transmission is G.N., but Morgan independent front suspension is cleverly incorporated, with lighter coil springs than standard. The engine is a 1,097 c.c. V-twin Blackburn, designed in 1932 by Robin Jackson. It has push-rod o.h.v. and is blown at 14 lbs. per square inch by a chain-driven Marshall supercharger set behind the crankcase. The fuel tank above the engine is of aerofoil shape to direct air on to the cylinders, and it feeds by gravity to an S.U. carburetter with Bowden throttle control. The driver sits ahead of the engine and has only a streamlined “coal-scuttle” for his legs ahead of him. The engine is amidships. It runs up to 6,000 r.p.m., and. cooling is something of a nightmare. Lubrication is by drip-feed. The frame is braced by a single cross-member and is very rigid, engine and clutch mountings acting as stress members. It is underslung at the back, and the fairly soft springs are wedged up. The Rudge front wheels carry 4.00″ x19″ tyres, and have Rudge brakes, cable operated, and the rear tyres are 6.00″ x16″. The rear axle is sturdier than the normal G.N. axle, and carries a single brake; all brakes have special Ferodo linings. A large rev. counter is set in a cowling right before the driver. The steering is rack and pinion from a pre-1914 Morgan. The “Freikaiserwagen” runs on straight dope and weighs under 6½ cwt. It has had some outstanding successes and does 50, 65, and 90 m.p.h. on the indirect gears, and crossed the line at Syston at 102 m.p.h. in top gear, still accelerating. The writer has often exceeded 100 m.p.h. in ordinary cars, but David Fry must experience very much accentuated emotions at this speed ill the nose of his very light vehicle!

Another very potent V-twin is Gerald Sumner’s Sumner-J.A.P. It started as the G.N. “Gehenna,” driven by Charles Martin and others. In 1935 E. C. Baragwanath designed a special push-rod 998 c.c. J.A.P. sprint engine, of which only three were built. This engine had duralumin-filled push-rods to obviate distortion, and a No. 8 Cozette compressor mounted behind the near side cylinder, blowing at 14 lbs. per square inch, and drawing from a special Amal carburetter, of which only two were made. The G.N. frame was extensively drilled, and fitted with this engine. During the winter of 1936 the chain transmission was replaced by a “2.3” Bugatti gearbox in a Brescia Bugatti shell, driving to a 2-litre Bugatti rear axle. The rear brakes are Bugatti and the clutch a Borg and Beck. A Special tubular front axle, with M.G. brakes, is sprung on quarter-elliptic springs, with radius arms below and friction shock-absorbers beside the leaves. The rear axle is sprung on quarter-elliptics. Doubtless this resulted in increased weight, permissible because the engine was now giving about 90 b.h.p. per litre; assuming the ratios were correct. The G.N. transmission has to be strengthened to transmit such power, and chain-breakages while changing speed become increasingly troublesome, but it does score in low weight, speed of change and rear axle adhesion, and in easy selection of useful ratios for different venues. Baragwanath raised the compression-ratio to 7.25 to 1 for one cylinder and 6.5 to 1 for the other, and 89 b.h.p. was realised at 5,300 r.p.m. In July, 1938, a decent body was fitted, with cowled nose, the oil-tank being so shaped as to direct air on to the cylinders. There is a 4½ gallon fuel tank in the tail. Remarkably noisy, the Sumner-J.A.P. can do about 112 m.p.h. and clocked 21.91 sees. at Lewes in 1938. It crashed rather badly at the last Prescott meeting of 1938.

Of more normal G.N. development is E. G. M. Wilkes’s Wilkes-G.N. Wilkes started racing a stripped G.N. known as “The Maggot” at Dancer’s End in 1932, but as this car—rumour says it was bought for 6/9!—was an almost standard i.o.e. touring G.N. it was not terribly impressive; it would possibly have been a better policy to have run it with road equipment and original body. In 1935 Wilkes acquired a “Mowgli ” engine from Sumner and, proceeded to overhaul both this engine and his original chassis. The rear springs were flattened and mounted on 2″ oak blocks, and new radius arms fitted. The front axle is tubular with Austin Seven brakes and king-pin assemblies welded to the tube, and rod brake operation. Later the front springs were flattened and placed above the frame. The engine was steadily developed, the valve timing being modified so that the inlet valves opened just before T.D.C., and two single-cylinder B.T.H. magnetos fitted. The inlet ports were welded up to 1.25″, tapering towards the valves, and the exhaust ports ground out. Two Amal Type 6 needle-jet single-float carburetters were fitted. To combat chain breakages, the transmission was rebuilt, and a new rear axle made up from a tubular shaft and M.G. brake components. The Wilkes-G.N. carries no body, but there is an oak facia before the driver and the car is beautifully turned out. Although not considered quite rapid enough for Shelsley, it has done extremely well at Dancer’s End, Beechwood and Poole. Extensions to the exhaust pipe improved acceleration at the last-named venue.

Another car now very highly developed is C. R. Instone’s G.N. “Martyr.” in 1929 Instone, along with two friends, acquired an early G.N. with push-rod engine. Throughout the winter much modification took place, twin magnetos firing two plugs per cylinder, and two carburetters being incorporated. Pat Thompson was the brainy gent, according to Rupert, and when his theory and science failed him he went on working— with a 3 lb. Brummagen screwdriver and an enormous crowbar, so that the rest of the team prayed a little something might be left as a souvenir. In 1931 two disasters occurred. Pat returned to Ireland, and the engine blew up most of its valve gear. A new J.A.P. push-rod engine was then installed, but, in spite of a modified clutch, it merely tied the frame in knots. At Shelsley it proved quite impossible to get out of third speed, and the entire climb had to be made on that ratio. The frame was accordingly braced up, and other mods. completed. Then the bevel-drive shaft and axle bent. So the entire transmission was overhauled, the shafts all being replaced in special materials, with the exception of the propeller shaft. The “Martyr”, now climbed Shelsley in 49.8 secs. although subsequent inspection showed that the prop. shaft had broken in the bell housing and was driving on the serration of its fracture only! A new clutch and propeller shaft were put in, and only the original side-members remained of the former G.N. A quite reasonable body was fitted, and wheelspin defeated by the use of Dunlop E.L.P. 5.75″ x16″ rear tyres. Later still, the engine was Zoller supercharged, and after overturning in practice at Shelsley last year, it took Rupert up in 48.0 secs., winning the supercharged 1,100 c.c. class. Even now we have not exhausted all the V-twin Specials. There are the Blackburn-engined Killick and the Carl-Mark J.A.P., both with engines behind the driver, and the converted three-wheeler Morgans of Lones (“Tiger-Cat”) and Barry ‘Woodall (“Chatterbox”), and a number of other G.N. developments.

Those V-twin specials that have been raced of recent times are likely to run against conventional racing-cars again after the war. Although they mostly date from the early G.N. days, their subsequent development has been a highly specialised business, on modern lines. It is hoped that the foregoing survey, if incomplete, will at least be of interest to those who seek individual descriptions of such cars, and that it may provide some applicable information to those who are working on V-twin sprint cars at the present time. To install a multi-cylinder engine in the simple chassis has been done times without number, and eventually was resorted to commercially by G.N. themselves. But it imposes great stress on the transmission if a useful power-gain is achieved, and if extra power in some abundance is not forthcoming, there is a loss on account of greater weight of block, head, crankshaft and cooling elements, etc.

So there is some basis for considering the air-cooled V-twin as a sprint job and for believing in its continued development after the war. It may well be that the research work on air-cooled aero-engines now in hand will especially focus the enthusiasts’ attention on the simple engine when speed trials return, and enable even more than 100 b.h.p., from 1,100 c.c. to be realised, with further reduction in weight, when, if adhesion problems can be lined up, the conventional racing job with lots of little boosted pots and shaft transmission will not look so modern.