Racing car evolution part II: 1909-1916

In the issue of March, 1942, we published a long article by Cecil Clutton dealing with racing car evolution between 1895 and 1908. This contribution was so very well received that it occurred to us to continue this account of the development of the racing car down the years. Articles of this nature involve an immense amount of thought and research, and the author of the original was quite unable to devote further time to the task. However, we have been offered, largely with his assistance and good will, further articles, each by a recognised authority for the period concerned, and have pleasure in presenting a second one herewith, in which Anthony S. Heal covers the years 1909 to 1916. Heal is, of course, well known not only as a collector of historic racing cars, some of them of the vintage covered in this article, but also as a very capable driver of such cars under modern competition conditions. The next article on evolution will deal with the 1919 to 1927 era, and after that the period 1928 to 1933 will be dealt with. Articles on the G.P. races of 1934 onwards were published in Motor Sport during the early part of the war and the voiturette races of 1935 to the outbreak of hostilities were also described. Although these latter articles were rather more a description of individual races than a study of evolution, nevertheless, sufficient technical data concerning the cars was included to carry on the story of racing car development. It is inevitable, with different writers covering separate periods of history, that the style adopted by Clutton cannot be maintained throughout, but the chief purpose of these articles, namely, to give an easily appreciated picture of technical development down the years, as emphasised by the happenings and results of classic International road races and other contests, is very well preserved. Additionally, this series is a very valuable quick reference to race results and to the specifications and outstanding design details of the better-known racing cars, from the very commencement of racing to the present time. We have only to mention that the next two articles are being contributed by Laurence Pomeroy, Technical Editor of "The Motor," and by Flying Officer John Scafe, R.A.F.V.R., to explain the pride we feel in announcing the continuance of this interesting and instructive series. – Ed. 

Like many other of man's activities, racing car design has experienced certain periods when it seemed that much further progress would not be possible and that a state of finality had been reached. But these times have always been followed by two or three seasons of intense development. Thus the introduction of the independently sprung Mercédès and Auto Union cars in 1934 came at a time when the then invincible monoposto Alfa-Romeo had reached the limit imposed by its chassis design. In 1924, when the 2-litre limit had already been in force for two years, the general adoption of the supercharger increased at one bound the power output of Grand Prix engines by 30 per cent.

Similarly during the period 1909-1910 the design of racing cars underwent a major change. Designers found that cars could be made smaller but yet faster, and the giant engines developed under the earlier Bore-Limitation regulations could be beaten by smaller faster running power units. Higher engine speed was substituted for long stroke. Lighter valve gear and semi-spherical combustion chambers were made possible by the use of twin overhead camshafts. Four valves per cylinder gave improved volumetric efficiency. Ball and roller bearings enabled shorter and stiffer crankshafts to be used. Dry sump lubrication was developed so that bearings were able to stand up to higher rotational speeds for prolonged periods. Shaft drive finally ousted chains for road racing and four-wheel brakes were used by more than one constructor in the 1914 Grand Prix. "Wind cutting" bodies and long tails showed that designers were beginning to be conscious of aerodynamics. With the reduction in engine size a great saving in weight was gained that made the cars easier to handle and lighter on tyres. The Rudge Whitworth detachable wheel emerged almost in the form we know it to-day to speed up wheel changes and, before "The Great War" put a stop to motor racing for some years, we see for the first time the tactical control of a Grand Prix team by a chef d'equipe signalling his orders from the pits.

The first three years of the period under review saw a decline in the popularity of racing and little progress was made. Trade depression and the recollection of the Mercédès win in the Grand Prix of the previous year caused the principal French firms to agree not to produce special Grand Prix cars in 1909. As a result, the A.C.F. had to abandon their Grand Prix. In the Coupe de "l'Auto" race for voiturettes, Guippone's successful Lion Peugeot (100 x 250 mm.) followed the lead set by Sizaire in previous years by using a single cylinder with a very long stroke. Three inlet valves, three exhaust valves and three plugs were provided. A two-cylinder Peugeot (80 x 192 mm.) was second. The highest placed car of anything like normal design was the four-cylinder (65 x 140-mm.) Hispano Suiza – a newcomer to racing. This voiturette race was the only event of any real importance that took place on the Continent in 1909.

Vauxhall and Singer were quietly developing small high-speed engines at Brooklands during 1910 and the 90 x 120-mm. Vauxhall achieved a speed of 100 m.p.h.

The supremacy of the single-cylinder engine for voiturette racing was again challenged by Hispano-Suiza in the 1910 Coupe de "l'Auto" with a long-stroke four-cylinder design (65 x 200 mm.), using a T-head. In spite of this apparent disadvantage, it was claimed that the engine gave 45 b.h.p. at 2,300 r.p.m. Shaft drive was used. Peugeot competed with a V-twin (80 x 280 mm.) with overhead inlet and horizontal exhaust valves. Final drive, as in the previous year, was by chain. The Hispano-Suiza won the race, scoring the first Coupe de "l'Auto" victory for a comparatively normal four-cylinder car. Hitherto the freak ultralong-stroke singles had been invincible.

Voiturette racing was the only form of competition which was serving any useful purpose. An attempt to revive the Grand Prix in 1911 was a complete farce from the development point of view, as it was won by a standard Type 61 Fiat which the concessionnaire had in stack as the result of a cancelled order. On the other hand, the 1911 Coupe de "l'Auto" produced some interesting designs. Bablot's winning Delage was quite a soundly planned machine. The cylinder dimensions (80 x 149 mm.) were not freakish and the horizontal valves, arranged symmetrically on each side of the head, were a step in the right direction towards the semi-spherical combustion chamber. The gearbox gave five speeds with a geared-up top gear, direct drive being on fourth speed. Final drive was by shaft. The Delage success was due to reliability. The Peugeots, which offered the main opposition, were probably quite as fast as the Delages and, in the hands of Georges Boillot, the Peugeot ace, possibly had a slight advantage. But despite the skill of the Peugeot team, Delage finished first, third and fourth. Boillot's car, which finished second, had four cylinders (78 x 156 mm.) arranged in the form of a narrow V. The \rakes were in the head, operated by a single overhead camshaft.

It began to be appreciated that the use of ultra-long-strokes was not the best way of getting maximum efficiency. Bore-Limitation regulations had been to some extent responsible for the concentration of designers' attention on long-stroke engines. The Brooklands Class records, being based on the R.A.C. horse-power formula, were also partly to blame. Thus, before the end of 1911 the Vauxhall Co. had taken records in both the 16-h.p. class and in the 21-h.p. class. The 15.9-h.p. Vauxhall (80 x 200 mm.) had a capacity of 4,021 c.c., while the car which ran in the 21-h.p. class had a smaller swept volume – 3,054 c.c. (90 x 120 mm.). The smaller capacity machine was the faster of the two. It was obviously an anomalous state of affairs, and in the following year the classes were revised on the basis of cubic capacity.

The year 1912 saw a renaissance of Grand Prix racing and the Gotterdammerung of the chain-driven monsters that had dominated the field until the decline which followed the 1908 race. In the Grand Prix de l'A.C.F. at Dieppe the "giant racers" made one final bid for supremacy, but in the course of the very strenuous two-day race over 956 miles they were worn down by the smaller and lighter Peugeot driven by Georges Boillot. It was Peugeot's first attempt at fullscale Grand Prix racing and their cars were designed by Ernest Henry. They were the first of a series of successful designs which were to become a basis for many rival manufacturers both in Europe and U.S.A. The four-cylinders (110 x 200 mm.) were cast en bloc. Two inlet and two exhaust valves per cylinder, inclined at 45 degrees, gave good volumetric efficiency, coupled with something approaching a semi-spherical combustion chamber. The valves were operated by two overhead camshafts, without any intermediate rocker gear. The cams bore direct on tappets, moving in guides, which in turn operated the valves. The camshafts were driven by a vertical shaft at the front of the engine. The crankshaft, was carried on three large ball bearings, which not only improved the mechanical efficiency of the engine but enabled a shorter and stiffer crankshaft to be used. Oil was fed under pressure to the main bearings and was carried thence by centrifugal force through oilways in the crank webs to the big-ends of the tubular connecting rods. A hand pump was used to lubricate the overhead camshafts. A disc clutch transmitted the power to the four-speed gearbox. Departing from earlier Peugeot practice, shaft drive was used instead of chains. Rudge detachable wheels were fitted. The engine and gearbox were carried in an elongated, U-shaped sub-frame mounted in the chassis on three ball joints. The idea was to relieve the engine and gearbox assembly from any chassis stresses which might possibly disturb the alignment of the crank and gear shafts. It was claimed that the engine gave 175 b.h.p. at 2,200 r.p.m. With an entirely unstreamlined body, consisting of two bucket seats and the classical bolster tank, Boillot was timed over a measured kilometre during the race at just on 100 m.p.h.

The principal opposition which these 7,6-litre Peugeots had to face was the "odd guard" as represented by F.I.A.T. and Lorraine Dietrich. Both had engines nearly twice the size of Boillot's car. F.I.A.T. had intended to use engines 140 x 220 mm., which had been specially designed for the race, but, owing to a strike at the factory, they had to content themselves with the 14-litre Savannah Grand Prix type of engine, which was 20 mm. shorter in the stroke but 10 mm. greater in bore. The cylinders were cast in pairs with vertical overhead valves operated hy a single overhead camshaft. Two plugs, two inlet and two exhaust valves were provided for each cylinder. The crankshaft was carried on three plain bearings and the drive was transmitted through a multi-plate clutch and four-speed gearbox. Final drive was by chains. Michelin detachable rims with 895 x 135 mm. tyres were fitted, and it is interesting to note that a tyre change could be carried out in 45 secs. with these rims. In spite of their enormous engines the Fiat's wheelbase was only 8 ft. 11 in., some 2 in. shorter than the Peugeots. The latter, however, had the advantage of lighter weight, scaling 17 1/2 cwt. as against Fiat's 22 1/2 cwt. The 14-litre engines developed 200 b.h.p. at approximately 1,700 r.p.m.

The battle lay between the huge, compact Fiat and the lighter French car. Each was handled with all the skill, dash and bravura that could be desired. Bruce Brown had a slight advantage on the straight (he covered the kilometre at 101.67 m.p.h., as compared with Boillot's 99.86 m.p.h.) and he skid-cornered by means of a special technique in which the hand brake and ignition switch were used.

Throughout the first day's racing the Fiat held the lead in spite of heavy rain, which must have been a greater handicap to the heavier car. On the second day Bruce Brown and Goux, on one of the other Peugeots, were both disqualified for re-fuelling on the course. De Palma (Fiat) was also eliminated for the same cause. Wagner's Fiat took up the challenge after his team mate had been eliminated, but was unable to wipe out the Peugeot's lead. Two laps from the end Boillot was delayed with a bent selector fork, but he straightened it, although he had to drive the last laps mostly in top gear. He won by more than 12 minutes, having averaged 68.45 m.p.h. for 956 miles. Wagner's Fiat was second at 67.32 m.p.h. Bruce Brown's Fiat put up the best time for the first day's racing and Wagner's Fiat was the fastest during the second day, which he started 27 minutes behind Boillot. By the finish the Peugeot's lead had been reduced to less than half.

During practice Boillot did a lap of the 47-mile course in 35 min. 55 secs. It is interesting to compare this with Nazzaro's fastest lap over the same course in 1907 in 38 min. 16 secs. and Mercédès best in 1908 in 36 min. 31 secs. The previous two-day Grand Prix was in 1906 at Le Mans, when Szisz on his 166 x 150-mm. Renault averaged 63 m.p.h. for 770 miles.

The Coupe de "l'Auto" for 3-litre cars was run concurrently with the Grand Prix and the three basically-standard sidevalve Sunbeams astonished everybody (including Louis Coatalen, I expect) by not only walking off with the first, second and third in the voiturette race, but, in addition, finishing third, fourth and fifth in the Grand Prix Race itself and by winning the Team Prize. It was the first victory for England in an important road race since Edge won the Gordon Bennett Cup in 1902.

Many interesting designs were entered for the 3-litre race, but unfortunately I cannot describe them all here. In addition to Sunbeam, Great Britain was represented by Calthorpe, Arrol Johnston, Vauxhall and Singer. All were adapted from standard productions. Singer's had ingeniously utilised a special head with rocker operated overhead inlet valves. The eight side valves were all used as exhaust valves. Vauxhall's had reduced their stroke to 118 mm. to come within the capacity limit. Calthorpe had lightened the reciprocating parts of their standard engine and claimed 68 b.h.p. at 3,000 r.p.m.

The "supreme Sunbeam" cars were basically the 12-16-h.p. (80 x 148.7 mm.) model, with slightly enlarged valves and stronger valve springs. The timing was nothing out of the ordinary; there was no overlap. A peculiar induction manifold, oval in shape, was litted with a special Claudel carburetter which seems to have borne some resemblance to the famous "triple diffuser "; 80 b.h.p. at 2,800 r.p.m. was claimed. The frontal area of the cars was reduced to the minimum and the narrow body was fitted with a "streamline" tail, of which the 32-gallon petrol tank formed part. Goodyear steel wheels were used.

The French cars consisted of the Lion Peugeot (78 x 156 mm.), a smaller and very light edition of the Grand Prix cars; Sizaire-Naudin (78 x 156 mm.), with horizontal valves (four per cylinder) and independent front suspension; Alcyon (85 x 132 mm.), with four overhead valves per cylinder, operated by pull rods and rockets; Th. Schneider, who had to use standard 80 x 140-mm. engines, as their special racing units were not ready in time for the race; and the two-stroke Cote.

The Sunbeam victory was largely due to their regularity. The value of Brooklands in developing the high-speed reliability of these cars cannot be overemphasised. The French constructors were undoubtedly at a great disadvantage in being unable to test their car adequately before the race. The 3-litre Lion Peugeots, in spite of their more advanced design, were slower than the basically standard Sunbeam and Vauxhall cars.

Resta's Sunbeam was the fastest 3-litre car over the kilometre, averaging 84.73 m.p.h. All the Sunbeams were faster than the other Coupe de "l'Auto" competitors, of whom the quickest were Champoiseau's Schneider (81.63 m.p.h.) and Percy Lambert's Vauxhall (80.4 m.p.h.).

Some Continental competitors made only half-hearted attempts at reducing the wind resistance of their bodywork. Others, particularly in the Grand Prix itself, had made no effort at all. Sunbeam, Vauxhall, Singer and Calthorpe, on the other hand, profiting by their Brooklands experience, all adopted some kind of "streamlining." All the British cars, with the exception of the Sunbeams, were fitted with V-fronted radiators.

It is interesting to notice the adoption, in these two Dieppe races, of various features which have subsequently become normal equipment of racing cars. Twin overhead camshafts were used successfully for the first time. Crankshafts carried on ball bearings were proved satisfactory in practice. Oil radiators were tried out by Sizaire-Naudin. Wire mesh stoneguards for protecting radiators were fitted to Peugeot, Fiat, Sunbeam and others. Revolution counters appeared on some cars, no doubt as a result of higher crankshaft speeds then coming into use. That their purpose was only slightly appreciated is shown by a comment in a contemporary journal, who offered the opinion that these instruments were "more for the instruction of the drivers during practice than for use during the race."

Peugeots scored another success at Le Mans later in the year when Goux won the Coupe de In Sarthe with one of the 7.6-litre Grand Prix cars, while Zuccarelli carried off the 3-litre race run at the same time. Early the following year Goux brought the Grand Prix Peugeot to Brooklands. Fitted with a cowling to the radiator and a long streamline tail, it started off by covering the half-mile at 108.5 m.p.h. and doing several laps at over 107 m.p.h. Finally, Goux took the 100 Miles and the One Hour Records at 106.2 m.p.h. The streamline body seems to have added about 15 per cent. to the car's speed.

Goux and Zuccarelli took two special Peugeots to Indianapolis for the 1913 500 Mile Race. The bore was reduced to 108 mm., but otherwise the cars were similar to the 1912 Grand Prix machines. The reduction of the bore was made to bring the cubic capacity within the 450 cub. in. limit. Goux won the race at 77 m.p.h., but Zuccarelli had to retire with bearing trouble.

The regulations for the 1913 Grand Prix, which was held at Amiens on a petrol allowance of just over 40 gallons for 570 miles (14 m.p.g.), forbade the use of streamline tails. The cars had to be fitted with cylindrical bolster tanks into which the allotted quantity of fuel was measured the day before the race. Tanks, filler caps, carburetters and petrol pipes were all plombes. Spare wheels might be carried behind the tank, but they must not be enclosed or covered over. Maximum and minimum weight limits of 1,100 kilog. (22 cwt.) and 800 kilog. (16 cwt.) were imposed just to make it a bit more complicated.

The Peugeot racing department, consisting of Goux, Zuccarelli and Ernest Henry, working under the direction of Georges Boillot, considered that somewhat smaller engines than those used in the 1912 Grand Prix would be necessary to eke out the petrol allowance for the 29 laps. Their 1913 engines differed slightly in other ways from those of the previous year. The cylinder dimensions were 100 x 180 mm. (5.6 litres) and the overhead camshafts were driven by gears instead of by a vertical shaft. The crankshaft was carried on three ball bearings as before, but the rear one was a double bearing. A rather unusual form of dry sump lubrication was employed. The oil was carried in a large tank under the driver's seat. Air pressure in the tank forced the oil forward, through six sight feeds on the dashboard, to the main bearings and other parts of the engine. A single pump returned the oil to the tank and maintained the air pressure. A combined ignition switch and oil tap was provided so that the sump should not be overfilled when the engine was switched off. The engine was set some inches further back in the sub-frame than previously. Large diameter brakes were fitted to the rear wheels and on the transmission. An ingenious worm and nut adjustment was provided so that the brakes could be taken up while the car was in motion. Top gear was 2.3 to 1. In other respects the 1913 Peugeot cars were very similar to those that had run at Dieppe. The appearance was much improved and the finish seems to have been much better. The narrow radiator and tapering bonnet, combined with the smooth undershield, gave the cars a very workmanlike appearance. Although only 138 b.h.p. at 2,150 r.p.m. was claimed (as compared with 175 b.h.p. for the 1912 7.6-litre engines), the maximum speed, as timed over a kilometre during the race, was 97.26 m.p.h. – only very little inferior to the speed of the larger cars the year before.

Georges Boillot won the Grand Prix at 72.2 m.p.h, and finished with some 5 gallons of petrol to spare. His consumption works out at approximately 16 m.p.g., which is very creditable. It was said that the Peugeots were designed to give 14 m.p.g. at full throttle. A water-jacket was provided on the Peugeot's induction pipe to assist vaporisation, but it was not found necessary to use it during the race. In 1907, when the Grand Prix was run on a fuel allowance, Nazzaro's Fiat won at an average speed of 70.5 m.p.h., but his petrol consumption was only 9.5 m.p.g. One interesting innovation was made by Peugeot's for this race which has since been adopted almost universally. The hub caps of their Rudge Whitworth wheels were fitted for the first time with the now familiar "ears" so that the locking rings could be removed by means of a hammer. Previously various kinds of spanners had been employed.

The chief opposition to the Peugeots came from the Delages and the Sunbeams. It was Louis Delage's first attempt at full-scale Grand Prix racing and the cars he presented at Amiens were very similar to the successful 3-litre car which carried off the Coupe de "l'Auto" in 1911. The four cylinders were in a monobloc casting, the bore and stroke being 105 x 180 mm., very similar dimensions to those of the Peugeot. Four valves per cylinder, placed horizontally, were actuated by push rods and rocking levers in the same manner as on the 1911 voiturettes. The plugs, two per cylinder, were placed in the head along the centre line of the engine. Two Bosch magnetos were provided, but only one was in action, the second was purely a stand-by. This pessimistic attitude towards the products of Robert Bosch A.G. was still current 12 years later when Sunbeams had a very similar arrangement of two magnetos on their 1925 Grand Prix cars. Delage supported his crankshaft on five ball bearings, whereas Peugeot only used three. Cooling was by thermosyphon. M. Clandel provided one of his then fashionable carburetters. A multidisc clutch transmitted the power to a five-speed gearbox (in the old Delage tradition) and thence by shaft to the back axle, which had a ratio of 2.4 to 1. (For the first time none of the cars in the Grand Prix had chain-drive.) The wheelbase of 9 ft. was slightly longer than the Peugeots, but the total weight was approximately the same.

Guyot's Delage pressed Boillot's Peugeot at some stages in the race and actually led the field from the tenth to the sixteenth laps. He finished fifth and Bablot (Deluge) fourth. Bablot actually put up the fastest lap of the race at 77 m.p.h. His speed over the kilometre, 93.2 m.p.h., was the second fastest recorded. Both Bablot and Guyot finished with nearly 6 gallons to spare.

Louis Coatalen, encouraged by the success of his Coupe de "l'Auto" cars at Dieppe, entered four six-cylinder 4 1/2-litre Sunbeams (80 x 150 mm.). Their design, as in the previous year, was based on standard productions. The cylinders, in two blocks of three, had side valves slightly inclined. The plugs were not screwed into the valve caps, as was then customary, but were placed in the centre of the head so that the plug bosses could he surrounded by the water-jacket. A Claudel carburetter supplied the mixture through a water-heated induction manifold. The crankshaft was carried on seven plain bearings. "High pressure forced lubrication" was effected at 40 lb./sq in. The engine was mounted in a sub-frame. Clutch and gearbox were of standard design, but the differential-less rear axle was of special design, very much lighter than the production job. The braking and driving stresses were taken through an I-section torque member attached to the bevel casing. The axle ratio was 2.5 to 1. Goodyear steel wheels with 880 x 120-mm. tyres were fitted as in the previous year. With a wheelbase of 9 ft. 10 in. the Sunbeams were longer and heavier than all the other competitors except the Italas; 100 b.h.p. at 2,750 r.p.m. gave a road speed of 103 m.p.h. Although any attempt at streamlining was prevented by the regulations, the frontal area was reduced as far as possible; the radiator was narrow and the bonnet tapered gradually into the scuttle. The spare wheels, carried behind the tank, were inclined, thus providing a slightly less abrupt finish to the bodywork than on the Peugeot and Delage cars, where they were mounted vertically. Provision was made for adjusting the brakes while the car was in motion and, to assist rapid warming up of the engine after the start, a roller blind, operated by a cord from the mechanic's seat, was fitted in front of the radiator. The petrol allowance encouraged the use of rather weak mixtures and the maintenance of engine temperature was therefore a matter of some importance. Chassagne and Resta finished third and sixth in the race and their speeds over the kilometre were almost identical – 88-89 m.p.h. Chassagne was only some 12 min. behind the winner at the end of the race.

Italas entered three large rotary-valve cars of 123 x 170 mm. (8,344 c.c.). They assumed that the large engine, somewhat underfed, would do better on the limited fuel allowance than the faster-running smaller engines. The engines were designed to run at 1,400 r.p.m. and gave 110 b.h.p. on the rather weak mixture that was necessary for the race. Without the fuel restriction it was claimed that 150 b.h.p. was available. The rotary valves, of which there were two – one for each pair of cylinders – were driven at one-quarter crankshaft speed from the timing shaft on the off side of the engine. The spindles of the valves were vertical and the valves themselves enclosed in the water-jackets. The cars suffered through lack of preparation, due to a strike at the factory, and when they came to be weighed they had great dificulty in finding enough removable parts to take off to come within the weight limit. Eventually by sawing off the exhaust pipe and removing the undershield, and even the hub caps, they scraped through. They were unlucky in the race: H.R. Pope retired in the first lap with bearing trouble, Nazzaro broke a spring on the thirteenth lap, and Moriondo, after turning over in a spectacular manner in front of the grand stands, managed to get the car back on its wheels again and pluckily continued until he was finally eliminated with a broken universal joint in the transmission.

Less than a month after the Grand Prix de l'A.C.F. at Amiens our old friends the Automobile Club de la Sarthe et l'Ouest organised another race at Le Mans, known as the "Grand Prix de France." The course was a good deal longer than the present-day "Circuit Permanent," being nearly 34 miles to the lap. Peugeot, Delage and Mercédès were the principal entries, but Peugeot did not start as the cars were said to have been sold to the U.S.A. The Delage cars were those which appeared at Amiens, but the Mercédès had not been seen before. Four were entered: Pilette drove a four-cylinder (140 x 150 mm. – 9.25 litres), Salzer and Lautenschlager handled "sixes" (105 x 140 mm. – 6.25 litres), while the fourth car (four-cylinder, 130 x 170 mm. – 8.9 litres) was entered by an amateur, Elskamp. The German cars made a good impression. It was Mercédès' first appearance in an International race since their victory in the 1908 Grand Prix. They were light in weight in spite of their large engines. The final drive was by chain, which seemed out of keeping with their otherwise up-to-date appearance. Pilette's car and the two six-cylinder machines were equipped with 893 x 135 mm. tyres on the rear wheels, and it was noticeable that the Mercédès had far less tyre trouble than the Delages, which had 880 x 120-mm. tyres. As on their 1908 cars, the chassis frame was upswept over both front and rear axles, the axle movement was damped by means of the traditional Mercédès opposed-cam-type of shock absorbers. The six-cylinder engines were said to have been originally designed as aero engines. Steel cylinders were cast in pairs. Only two valves per cylinder were employed. They were operated through exposed rockers by a single overhead camshaft; 90 b.h.p. was developed at 1,400 r.p.m.

Bablot and Guyot took first and second places with their Grand Prix Delages after a most exciting race. Pilette was third with the four-cylinder Mercédès and Salzer fourth on one of the "sixes." Duray (Deluge) just beat Lautenschlager on the second Mercédès "six" for fifth position. Bablot put up the fastest lap at 85 m.p.h. and averaged 76.8 m.p.h. for the race.

The 1913 Coupe de "l'Auto" was run at Boulogne over a twisty, hilly course, which had to be covered 12 times. The total distance was 387 miles. Engine capacity was limited, as previously, to 3 litres and maximum weight allowed was 900 kilog. (approximately 17 1/2 cwt.). The chief contestants were Peugeot, Sunbeam and Vauxhall. Peugeot had produced three new 3-litre cars (78 x 156 mm.), which were sealed-down versions of their Grand Prix cars. The design of engine and chassis was on the now familiar lines and the cars weighed over 1 cwt. lighter than any of their competitors.

Sunbeams had developed their 1912 cars, but they were still basically the same. Larger bearings, lighter reciprocating parts, higher oil pressure and 5 to 1 compression resulted in higher power output ) • 87 b.h.p. at 3,000 r.p.m. was claimed and the cars were said to be 6 m.p.h. faster. The differential-less back axle, as used on the Grand Prix racers, replaced the standard axle used in 1912. Radiators were covered with "wind cutting" cowls and the streamline tails of the successful Dieppe machines were retained.

The Vauxhalls, too, were much like the 1912 cars, with detail modifications to increase power. With 5 to 1 compression the 90 x 118-mm. engines developed 70 b.h.p. Larger valves and careful porting made for greater volumetric efficiency. Special "Zephyr" lightweight pistons with two rings and the minimum amount of skirt were employed. The crankshaft was 2 in. in diameter. Engine and-gearbox were set further back in the frame to get as much weight as possible on the rear axle, but this method of weight distribution made the cars difficult to hold on the bumpy course.

In the race the Peugeots were easily the fastest cars, Boillot and Goux finishing first and second at 63 and 62 m.p.h., respectively. Guinness with the Sunbeam finished close behind Goux, while Hancock (Vauxhall) canie in fourth. Guinness, who actually lay second from the seventh to the tenth lap, ran the whole distance non-stop. Resta and Chassagne's Sunbeams both retired with cracked rear axle casings. Whether this was in any way connected with the absence of a differential it is hard to say, for the Grand Prix cars did not experience any trouble.

Shortly after the Coupe de "l'Auto," Goux brought his 3-litre Peugeot to Brooklands. It had been fitted with a single-seater body and streamline tail. He covered the half-mile at 106.19 m.p.h. and ten laps (standing start) at 100.68 m.p.h. It is interesting to compare these speeds with those put up by the 1912 7.9-litre Grand Prix car which Goux brought over to the Track earlier in the year.

The Grand Prix and Coupe de "I'Auto" races of 1913 served to show that the "souped up" side-valve touring engines could no longer compete with the specially designed overhead-valve racing jobs. Mr. Coatalen and Mr. Pomeroy were quick to grasp this and set about designing over head camshaft units for 1914. The lessons learnt by Mercédès with their experimental cars at Le Mans were fully digested at Untertürkheim, and they set to work with real determination to produce a design with the avowed intention of winning the 1914 Grand Prix de l'A.C.F.

The R.A.C. revived the Tourist Trophy Race in 1914 after a lapse of six years. Just to be different the Club decided on a cubic capacity limit of 3,310 c.c. No explanation appears to have been given as to how this curious figure was arrived at. Why the organisers could not content themselves with 3 litres, which was the commonly accepted capacity limit for all important voiturette races, I cannot imagine. A minimum weight of 21 1/2 cwt. was imposed and it was stipulated, for some obscure reason, that cars should not lose more than 50 lb. in weight during the race. The course was the Isle of Man "Mountain" circuit, 37 1/2 miles round. Eight laps had to be covered on each of two consecutive days, the racers being impounded in an official enclosure overnight.

Several English manufacturers produced special overhead-valve racing engines for the event. Sunbeam, Humber and Vauxhall presented 16-valve engines with twin overhead camshafts. The first two firms had obviously been considerably influenced by Peugeot's Coupe de "I'Auto" machines. Straker Squire brought out a special engine with a single overhead cam shaft, but it was not used in the race as their standard side-valve unit was just as fast. Crossley, D.F.P., Hudson, S.A.V.A. and Star were all basically standard designs slightly modified for racing. Minerva entered a team of Knight sleeve-valve-engined cars.

The three Humber cars were designed by F. Burgess,, who later collaborated in producing the 3-litre Bentley. The cylinders (82 x 156 mm. = 3,295 c.c.) were cast en bloc. The two overhead camshafts were driven by gears from the front. The exhaust ports were on the off side, a rather unusual arrangement. Tubular connecting rods were employed. The engine was mounted in a sub-frame hung below the level of the chassis side members. The gearbox, however, was attached to the central cross members, so that some advantage of the sub-frame seems to have been lost. The drive was taken through the rear springs, which were slung below the axle; 100 b.h.p. at 3,200 r.p.m. was developed, but the designer claimed later that 113 b.h.p. at 3,900 r.p.m. was attained.

Mr. Coatalen, after his side-valve 3-litre cars had been defeated by the overhead-valve Peugeots in the Coupe de "l'Auto" race, decided that imitation would be the sincerest form of flattery. One of the 3-litre Peugeot racers duly found its way to Wolverhampton, and the 1914 T.T. Sunbeams bore more than a family resemblance to Ernest Henry's voiturette.

The late Mr. L.H. Pomeroy was more original and enterprising and produced some very interesting cars (90 x 130-mm. – 3,307 c.c.) which had a much lower stroke-to-bore ratio than most of the other competitors. Two overhead camshafts were driven by a worm wheel from a vertical shaft at the front of the engine. The 16 valves, containing only a slight dihederal angle, were mounted almost vertically. They were operated through rockers. The magneto was driven off the rear end of the inlet camshaft. The monobloc cylinder casting allowed ample water spaces around each cylinder barrel. "Zephyr" steel pistons were fitted as on the Coupe de " l'Auto " cars of the previous year. The crankshaft was counter-balanced, the balance weights being fitted into slots machined in extensions of the crank-webs and riveted in position. Higginson's car shed one of these weights during practice, which made rather a mess of things, but by dint of a lot of all night working the car was brought to the starting line. Crankcase breathing was achieved by means of small ship's ventilators, one facing forward, the other backward, projecting through the top of the bonnet; 90 b.h.p. at 3,600 r.p.m. was claimed. The cone clutch had the lining mounted on the flywheel half, to reduce the weight of the spinning member. A four-speed gearbox was bolted to the front end of the torque tube with a spherical joint, enclosing a universal joint, between the clutch and the gearbox. This unusual arrangement, the object of which I have failed to appreciate, entailed enclosing the transmission brake within an enlargement of the torque tube. Cantilever rear springs were fitted and the rear axle was provided with large ribbed brake drums, quite a departure from the usual Vauxhall tradition. The front springs were passed through slots in the front axle, a la. Bugatti.

The whole design of these cars bristled with interesting and unorthodox features. A great deal of attention was devoted to saving unnecessary weight and, like the Sunbeams, ballast had to be added to bring them up to the minimum weight. Unfortunately, the Vauxhalls suffered from lack of preparation as they were not ready in sufficient time before the race. Sleeve valves have not been seen very frequently in racing, although Argyll and later Panhard-Levasseur and Voisin achieved quite a measure of success and managed to break a number of records using engines of this type. Minerva's produced three 90 x 130-mm. Knight engined cars for the 1914 T.T., which ran with no little success. The inlet ports were increased in size compared with the standard production. To achieve this increase in size it was necessary to reduce the dimensions of the exhaust ports. Large supplementary exhaust ports were, however, provided on the off side near the bottom of the sleeves. These ports were opened 60 degrees before bottom dead centre at the end of each power stroke. Two motion shafts were fitted on opposite sides of the engine; one operated the inner and the other the outer sleeves. Great care was taken to ensure adequate lubrication of the sleeves. The cylinder bore was enlarged at one point, a horizontal annular groove, 1 1/2 in. wide by 0.02 in. deep, being machined in the cylinder wall. Oil was fed under pressure into the groove on one side of the engine and was drawn off from the opposite side of the engine by a scavenge pump. Thus oil was positively circulated around the sleeves at all speeds. A third oil pump attended to the main and big-end bearings and an oil cooler was mounted behind the water radiator. The engines ran up to 3,300 r.p.m. and had twin carburetters.

As might be imagined, reliability of the sleeve valve was achieved under racing conditions only at the expense of considerable oil consumption. During the practice the dense smoke screen put up by the Minerva cars on certain parts of the course was a great inconvenience and a potential danger to other faster competitors. As a result there was talk of not allowing the Belgian team to start, but after a test the R.A.C. decided to let them compete in the race on the understanding that the mechanics kept a constant look-out for any car wanting to pass. One interesting point was noticed at the start of the second day's racing. Like most competitors the Minervas were pushed from the starting grid to their depot for replenishment before setting off. They had only covered 300 miles during the first day, but they absorbed 10 gallons of oil without any difficulty!

The Sunbeams were the fastest cars in the race. K. Lee Guinness did one lap in practice in 36 min. 45 secs., an average of over 60 m.p.h. Resta was eliminated early in the race, but at the end of the first day K. Lee Guinness led the field at an average speed of 54.7 m.p.h. His brother Algernon lay second with a sister car, only 3 min. behind. Porporato (Minerva) was third. During the second day Algernon Guinness retired with the rear universal joint damaged, but "K.L.G." upheld the Wolverhampton firm's colours, finishing 20 min. ahead of the second man, Riecken (Minerva), at an average of 56.44 m.p.h. for 600 miles. Minervas occupied second, third and fifth places, thus winning the team prize. Witchell's Straker Squire was fourth and W.O. Bentley on the little D.F.P. finished sixth.

The rules for the Grand Prix de l'A.C.F. held at Lyons a month before the Outbreak Of the last "Great War" permitted a maximum capacity of 4 1/2 litres and a maximum weight of 21 cwt. The race itself was certainly one of the high spots in the history of motor racing. Forty entries, representing England, France, Germany, Italy, Belgium and Switzerland, made the race a really International competition and the design of the competing cars was of great technical interest. Designers all agreed, with one exception, that overhead valves were necessary to get the utmost power from engines of limited capacity. Of the 13 marques entered, five used twin overhead camshafts, six were able to manage with one, while the two Piccard-Pictets from Switzerland had single sleeve-valve engines built under Argyll licence. The Aquila-Italiana was the only six-cylinder car, the rest all having four cylinders. Except for two Italian entries (F.I.A.T. and Aquila-Italiana) and the sleeve-valve Swiss entries, every firm followed the Peugeot tradition of using four valves per cylinder. Plain crankshaft bearings were used by a slight majority, including the winning Mercédès, but ball bearings were chosen by a considerable number of manufacturers, including Peugeot, Delage and Sunbeam. Th. Schneider was alone in selecting roller bearings. The most important innovation made by several constructors was the use of four-wheel brakes, for the first time in an important road race, by Peugeot, Delage, F.I.A.T. and Piccard-Pictet. The use of brakes on the front wheels was no doubt influenced by the very twisty nature of the course, which put a premium not only on powers of acceleration, but also on deceleration. F.I.A.T. and Peugeot were alone in making anything like a serious attempt at streamlined bodywork.

Mercédès, who had the avowed intention of winning the Grand Prix after an absence of six years, had profited by their preliminary canter in the Le Mans race of 1913. The Ilve cars which appeared at Lyons, although quite fresh designs, showed that certain lessons had been learnt from the machines of the previous' year and from their experience of making aeroplane engines. The four-cylinder engine (93 x 165 mm.) had separate steel cylinders with welded steel jackets in the aero-engine manner. A single overhead camshaft, driven by a vertical shaft at the rear, actuated four valves per cylinder through rockers. The valves were inclined at 60 degrees, the valve springs and the ends of the rockers being outside the camshaft casing. Four sparking plug holes were provided, but only three plugs were used in the race, two on the inlet and one on the exhaust side. Two magnetos, one a twin-spark machine, were driven by a cross shaft at the base of the camshaft drive. A single Mercédès carburetter supplied the mixture. The crankshaft was carried on five plain bearings. The lubrication was very completely thought out. The oil pump, which was an affair like a small multi-cylinder steam engine, not only distributed oil to the main bearings, the overhead camshaft and the cylinder walls, but it also drew a fresh supply of lubricant from the reserve tank, thus maintaining the level in the sump. In addition, a small pump, operated by the mechanic, could be used to force an additional supply of cool oil direct from the oil tank to the main bearings, camshaft and cylinder walls. A screw-down lubricator in the cockpit could be used to force grease to the water pump while the car was in motion. The drive was taken via a typical Mercédès double-cone clutch to a fourspeed gearbox which was supported from an X-shaped cross-member (reminding one of the much-advertised "cruciform construction"). The rear axle was unusual in that two crown wheels and two pinions were used. The drive was taken through a stout tapering torque tube which ended in a large spherical joint just behind the gearbox. The universal joint was lubricated from the reserve tank. The forward end of the chassis was upswept over the front axle. The usual Mercédès pattern opposed-cam type of shock absorbers were fitted both front and back.

Mercédès claimed that the engines developed 115 b.h.p. at 3,000 r.p.m. and this does not seem to be an extravagant claim when it is borne in mind that the cars were capable, without the aid of streamlined tails, of over 110 m.p.h. It seems probable that 3,500 r.p.m. was achieved at maximum speed. Of the five cars entered, three driven by Lautenschlager, Wagner and Sulzer took the first three places. Sailer in his duel with Boillot's Peugeot put up the fastest lap, while the fifth car, driven by Pilette, was crashed.

The general arrangement of the Opel engine was rather similar to Mercédès, although the cylinders were cast en bloc. The dimensions were 94 x 160 mm., measurements that were also used by Alda, Deluge, Sunbeam and Nazzarro. The Opel cars were higher off the ground than the low-hung Mercédès and weighed 18 cwt., some 2 cwt. less than their compatriots.

Peugeots naturally offered the most serious competition to the German cars, although the Deluges had proved very fast in practice. The engine was of the usual Henry type, measuring 92 x 169 mm. – the longest stroke of any engine used in this race. The most obvious departure from previous practice was the adoption of four-wheel brakes; 15 in. drums were fitted front and back, the former being operated by the foot pedal and the latter by the hand lever. Both axles were underslung and their movement was damped by the traditional Peugeot friction-band shock absorbers. A very neat streamlined tail enclosed the spare wheel, which was mounted fore and aft, instead of at right angles, behind the petrol tank. The Peugeots lacked the speed of their German rivals, but Boillot, by a display of heroics and full use of his superior braking, was able to hold either first or second position right up to the last lap but one, when he had to retire with a valve through a piston head.

The quest for higher r.p.m. decided Louis Deluge not only to do away with the comparatively heavy rocker gear and push rods used on his 1913 cars and to follow the general trend by using two overhead camshafts, but, in addition, to experiment with mechanically closed valves, thus eliminating the ordinary valve springs. Th. Schneider also had positive valve operation, but it worked in a rather different manner. The Delages had a monobloc cylinder casting (94 x 160 mm.) with two overhead camshafts driven by an enclosed vertical shaft at the front of the engine. Each camshaft bore four cams, which were each surrounded by a steel ring. These rings were not truly circular, but D shaped. Each formed a tappet which the cams could alternately press downwards to open the valves or upwards to shut them. Each tappet operated two valves (there were four valves per cylinder) through a T-shaped yoke. The cams lifted the valves against their seatings and held them there through small buffer springs fitted to the yokes. The arrangement was apparently unreliable and, although Deluge had promised well during the practice period, only one managed to finish the race. The balanced crankshaft was built up in four parts and was carried by five ball bearings. A single Claudel carburetter was fitted, although experiments were made with one for each cylinder. The gearbox gave five speeds, as previously, but on this occasion direct drive was on third. The first overdrive was for use on the straight and the highest ratio of all was brought into play when going downhill. Perrot-type brakes were fitted to the front axle. Large drums the same size as those at the rear were used. All four brakes were operated together by the pedal.

Sunbeams presented three cars very similar to their successful T.T. design. The engine was, of course, larger (94 x 160 mm.) and the valve gear differed slightly. Two Claudel carburetters were fitted. The wheelbase was increased to just over 9 ft., which no doubt served to improve the car's directional stability. Although very little practice had been possible on the rather tricky circuit, the Sunbeams showed up fairly well in the race. Either Resta, Guinness or Chassagne was among the first seven on every lap and Resta finished fifth behind Goux's Peugeot. The Sunbeams were actually faster than the Peugeots, but the latter had the advantage of four-wheel braking, which was a considerable gain on the Lyon circuit.

Vauxhalls also produced larger editions (101 x 140 mm.) of their T.T. cars, but lack of preparation prevented them from showing their true capabilities. The Pomeroy design had the smallest stroke-bore ratio in the race.

The Mercédès were undoubtedly as fast as any of their competitors, possibly faster, and by that meticulous attention to detail which we have witnessed both in racing and warfare during recent years, they made certain of their victory. The cars were directed as a team and controlled by the team manager in the pit, which was quite a new idea.

Within a month of the Lyon race Europe was occupied with war and motor racing became a thing of the past for most people. The Sunbeam Company, however, managed to construct two six-cylinder racing cars and ship them to America for the 1916 Indianapolis race. How they managed to do this without awkward questions being asked by the Air Ministry, for whom they were making aero engines, I cannot imagine. The six cylinders (80 x 156 mm. – 4,705 c.c.) were cast en bloc. The sides of water-jackets were formed by aluminium plates. Two overhead camshafts operated four valves per cylinder through rocking levers. The crankshaft was carried on ball bearings and dry sump lubrication was effected by two gear-type oil pumps. Two Claudel carburetters were employed. The rest of the chassis followed the same layout as that of the Grand Prix cars. Christiaens finished fourth at Indianapolis behind two Peugeots and a Duesenberg. He drove in a number of races, but Resta's 1914 Peugeot always had the advantage. The best the Sunbeam could manage was a second (behind the Peugeot) on the Sheepshead Bay Track, at an average speed of 104 m.p.h.

With the entry of the U.S.A. into the "Great War" racing ceased in America, as it had in Europe, and thus came to a close one of the most interesting periods of development in the history of motor racing. Although the cubic capacity of Grand Prix cars had been reduced from 14 to 4 1/2 litres, between 1912 and 1914, the smaller cars were 10 per cent. faster in maximum speed and, with their lighter weight and improved braking, their average speed on a twisty course was immeasurably superior.