Vintage Sports and Racing Salmsons



Vintage Sports and Racing Salmsons

by John McLellan

SALMSON PUBLICITY PORTRAIT of ti 1928129 production Grand Pri:s Model with metal body and balloon tyres.

IT is easy E0 forget that the vintage period in motoring lasted only a decade and that many of the now legendary marques were introduced, flourished, and declined during a shorter period than the production run of some of today’s popular models. Although their keenest protagonist would never claim that Salmsons were the stuff of legends, they conformed to this pattern, for their entry in any important competition from 1921 to 5917 was usually viewed with some despondency by other competitors in the same class; but between 1927 and 1930 the twin bogies of the vintage manufacturer, high weight and development costs, made themselves felt and Sihnsons slipped into mediocrity. The make came into being when M. Armand Bovier, a Frenchman domiciled in England, suggested that production of a cyclecar would be a good way for the old-established Societ6 des Moteurs Salmson to absorb some of the profits they had made from the construction of aero engines for the Allies during the Great War. Bovier had been educated at L’ecolc Industrielle and had served an apprenticeship with Lacoste. After some years spent working in France and Russia he came to this country for the Tli. Schneider company. A keen racing enthusiast, he had competed at Brooklands just before the war with a 5A-litre Schneider. During the war he won the Croix de C”hterre and was invalided from the Services. He returned to assist the war effort by building Rhone et Gnome aero engines and doing contract work for A.E.C. at his works in Kilburn. The S.M.S. had been producing rotary and radial aero engines of individual design for several years. At the end of the war their plant at Billaneourt covered an area of 15O acres and 10,000 work-people were employed there. Hosier knew that S.M.S. were ionising for a car to start them off as motor manufacturers. Ile was convinced that the light-car class would he the most popular in the years tO cOme and realised that a cyclecar, simple and straightforward to Make, would be the best machine for an inexperienced organisation to copy. The G.N. was the car which appealed most to him, both on the strength of a successful racing career before the war, and for its simple and sturdy design. It had proved itself to be comparatively practical and controllable as well as fast, outstanding virtues in a member of a group of vehicles which never became famous for these qualities. The G.N. chassis was to form the basis of innumerable “specials” and at

least three other marques. Fortunately Messrs. Godfrey and Nash were agreeable to the manufacture of their car under licence in France. So the first Salmson was announced in March 192t. A rakish little machine, it was a straightforward copy of the G.-N. ” Standard ” model. There was a straight channel section frame suspended from four 1-elliptic springs. The engine was the famous i.o.e. vee-twin eleven-hundred driving from the front of the chassis via the clutch and a long shaft to a bevel-driven layshaft at the rear, which in turn drove the solid bar rear axle by any one of’ four sliding-dog-engaged chains. Brakes were fitted on the rear axle only and speed selection was by a right-hand lever. Steering was very direct by a crown-wheel and pinion to the tubular front axle. The whole ensemble was really light and owed what per

formance it had more to this rather than to the modest engine output.

In fact, during a period when most models became regrettably heavier year by year, sporting models of Salmson and Frazer Nash, both based on the U.N., resisted this tendency better than most, and remained comparative lightweights. This was an important factor in the:success of the :Salmson racing cars.

The Salmson U.N. was made in fairly substantial numbers during 1921 and the ensuing years, and a total of 3i000 were built. They were used not only by sporting owners, but by people who, wanting a cheap economical car, were willing to brave the weather in these spartan little machines. They were even used by the Paris police, one can only guess for what purpose! The standard model was a narrow two-seater with slightly swept wings and long fiat running-boards. A hood was supplied and the spare wheel was carried on the short curved tail. An oil tank was carried on the driver’s running-board, with the hand primer just beside the gearlever.

This model was produced in ” Type Sport ” form, with an even sketchier body than the ” Standard,’ and prototypes were entered in a few events during 1920. For 1921 the team embarked on a full programme of competitions and races. The cars were used in two forms; for minor events the “‘Type Sport” body was fitted, for more important speed events and races a neat and attractive racing shell was devised. This was a slim two-seater with the mechanic’s seat considerably staggered back from the driver’s. The two cylinders of the engine could be Seen protruding from the slender pointed nose of the body, which swept out to the full width of the chassis at the scuttle and finished in a long high-slabsided tail. The two spare wheels were fitted transversely behind the seat squabs and there was a small wire mesh aero-screen for the driver. The engines fitted to these cars were mildly tuned, carefully assembled units with slightly inclined overhead valves and hairpin valve springs. Flared wings were provided at the front whilst at the rear two horizontal strips sufficed. These cars were registered for road use, and took part in both sports and racing events. They were driven to and, in most cases, from meetings. At the end of the season the team were declared ” Champion de France Tourisme 192 t .” However, it was intended that the cyclecars should be superseded by cars built more on light-car lines as soon as possible, and in the summer of 1921 examples of the new cars, to be known as the 7 c.v., were tested. Two engines were designed for the new car; one was an advanced 4-cylinder twin-o.h.c. unit, which was built only as a prototype during 1921, the second was the famous 4-push-rod engine, a simple design intended for immediate production. The chassis layout of the cyclecar was modified, and the new o.1). a-cylinder engine was designed by M. Petit, who was to he responsible for the Salmson designs for very many years, and who was still employed by the company when it became part of the Renault organisation in the middle ‘fifties. The engine dimensions were of 62 90 mm., (5,087 c.c.). The cylinder block was of cast-iron with a detachable head, mounted on an aluminium crank

case in which the crankshaft was carried in two plain bronze bearings. The valves were inclined, and the series of eight were operated by four rods and four rocker arms. For the exhaust valve the spring was on the stem in the usual manner, but the inlet was operated by the same rocker and had two small coil springs, anchored at their ends to a cross-pin on the rocker, and to lugs on the cylinder head. Each cam was recessed and it was on the down movement of the push-rod that the inlet was opened. Derived from Salmson aero engine practice, the system worked well enough provided that engine speeds were kept fairly low. The upper part of the engine was lubricated by a pump driven from the front of the crankshaft and the lower half of the engine was splash lubricated. The magneto, of Salmson make, was opposite the dynamo, driven from a cross-shaft at the front of the engine. A Ferodo-faced cone clutch drove the unit gearbox. A neat and compact 3-speed design with short shafts and quite large diameter plain bearings, the ratios were 4 to r, 61 to r, and r6 to 1. Simplicity and lightness was the keynote of the rear axle, of straightforward design, with a steel torque tube and banjo casing to which were attached light-alloy axle tubes and brake backplates. The straight-cut crown-wheel and pinion dispensed with a differential and the rear-axle shaft was a solid bar. Brakes were on the rear axle only, operated by cables led over pulleys and down inside the radius-arms to the brake levers. The drums were cast-iron. The remainder of the chassis retained the G.N. layout. Light wire wheels, fitted with 700 < 8o tyres, were retained by a primitive centre-locking ring. All Salmsons using this fitment were prone to shed their wheels and eventually a more satisfactory method was devised. The chassis weighed 5 cwt. and the complete car weighed 8 cwt. in standard two-seater form, or 9 cwt. 2 qr. with de luxe equipment and a dickey seat.

There was still a strong family resemblance to the G.N. in the appearance of the model; the spidery effect was enhanced by the well-set-back flat radiator with its “X ” motif, representing the cross-bracing of the front wings of the racing models; and by the tiny-section straight-sided tyres. It was available either as a twoseater having a small dickey seat with well-padded backrest, or as a ” Type Sport ” with the skiff tail which had such a vogue in French light sports coachwork circles in the early ‘twenties. The wheelbase was 8 ft. 6 in., and the tyre section was 7007: go. The spare wheel was now carried set into the driver’s running-board and the wings were endowed with adequate valances. A cape-cart hood with canvas sidescreens was supplied, but there was no windscreen wiper; in bad weather the intrepid driver had to peer through the gap between the windscreen, low and sloping, and the hood rail. Standard paint finish was blue, set off by a polished aluminium bonnet, the whole effect completed by the black wings.

Production versions of these models were announced for 1922. By the spring development work on the 7 c.v. model was complete and the factory was ready to swing into full production, intending to turn out 150 cars a week in March, rising to 20011 week during April.

In this country the company embarked on a full sales campaign from the premises in Buckingham Palace Road, and later in Motcomb Place. Cars were loaned for road-test and plenty of descriptions found their way into the pages of the technical Press, enthusiastic owners entered their cars for trials and long-distance runs, and in the Scottish Six Days’ Trial a gold medal was won by D. Brown. There was some head-shaking over this entry prior to the event. The feeling was that the gear ratios were not really suited to this kind of motoring. Afterwards The Autocar commented that they had doubted the gear ratios, but now, considering that the engine gave 18 b.h.p. and the equipped car weighed 9 cwt., the ratios were better than most; 2 h.p. per cwt. was an excellent ratio for a little touring car able to accept a top gear ratio of 4 to for ordinary roads, besides being able to climb normal slopes really fast on a second of q to t, whilst a 16 to 1 bottom was a genuine dreadnought for so light a machine with such a good engine. They applauded the makers for sending so standard a gar. ” A 2.5% drop in gear ratios would have presented an impressive spectacle, but would have been less honest to their customers.” The 7 c.v. was known in this country as the 9.5 h.p. model and was tested by The Antocar in August, t922. The car used was the ” English “-bodied 2-seater de luxe with dickey. In the course of an 8o-mile run out to I3ox LIM they took the car to Brooklands, where it was timed at 50.7 m.p.h. for the measured mile. The test hill was climbed from a standing start at an average speed of to.26 m.p.h. Acceleration figures were 10 to 30 m.p.h. in 12; sec. in top gear, whilst the same test took q sec. through the gears. Two people were carried in all these tests. Box Hill was climbed mainly in top and second, first being used only on the hairpins. Later the car demonstrated it had ample power on Pebblecombe Hill, climbing in first with a good deal in reserve. During all this testing the overall fuel consumption worked out at

The testers found the car comfortable, with good cushions and quite soft springs. There was ample leg room for the tallest driver. The positions of the gear and brake levers were criticised, it being found that the knob of the gear-lever was first on one side and then on the other of the similarly shaped brake lever, Salmsons never really sorted this out and few Salmson owners with Standard levers have avoided trying to change gear with the handbrake from time to time. The dickey seat was a little cramped. but offered excellent luggage accommodation.

Their conclusion was that the little Salmson offered a road performance which was really impressive and exhilarating. Indeed, at the price there were few cars in the class to touch it.

The owners of the 9.5 h.p. cars were often enthusiastic. One Owner writing in April 1923 of his similar 1922 machine says that he had driven his car for a year, covering to,000 miles. Durability of the engine, body and chassis was excellent, he had decarbonised the engine twice and all the hearings were tight on each occasion. Ile had made three return trips of 4.40 miles from Yorkshire to South Wales within a fortnight, during which time the bonnet was lifted only once, due to a sooted-up plug. The return journey could be covered comfortably in a day, and the longest non-stop run was 165 miles. Tune was well maintained provided that the valve clearances were checked carefully every 2,000 miles, when the crankcase was drained as recommended. The cruising speed was 30/33 m.p.h., and at 40 m.p.h. the car held the road like a large car. Overall petrol consumption worked out at 40 m.p.g. The suspension was very satisfactory when loaded, but rather harsh with only the driver on board, and springs gaiters were considered a worthwhile extra. Average life of a tyre was 7,000 miles.

It was against this background of successful acceptance that development of a larger car was proceeding; a car that would be able to carry the larger bodies, offering better and more comfortable accommodation, that were becoming Popular even with the owners of small-engined cars; And the opening of the racing season in 1922 saw the announcement of a remarkable team of Salmson racing cars.

In their original form the chassis and engine layout of the racers was identical to the new 10/15 model, which was to be announced at the 1922 Paris Salon. The high performance Of which the cars were capable was still due more to their light weight and clean profile than to the power given by the engine. But the line, new, little twin o.h.c. unit had a great potential for development, and allowed the team to keep pace with its competitors for some year’s.

The basis of the model was a long, slender, flexible chassis of tapered channel section which gave a wheelbase of 8 ft. 7 in. and track of 3 ft. 9 in. At various times there were half-hearted attempts to stiffen the chassis by boxing the side members on sprite cars, but a more typical solution was to fill the chassis with ash strips to reduce vibration and guard against frame fractures. Suspension was by I-elliptics at the rear, with 1-elliptics at the front; these were pivoted at their front ends and slid in primitive trunnions at the rear. Shock-absorbers were not Used, but rubber straps guarded against excessive spring movement. There was remarkably little chassis development on this model over the years, but with the addition of front-wheel brakes and with the high speeds achieved the short-comings, of the suspension was realised and the springs were carefully bound with cord.

A light front axle was used, consisting of a straight tube with forged end pieces sweated and pinned into each end. At first there were no front-wheel brakes. The rear axle was identical to the smaller to h.p. car, with the same small cable brakes.

In 1927 M. Petit said that he considered that a special racing or competition car would be in the hands of the ordinary motorist within two years. 111.1922 this certainly was true, for if is difficult to say which came first, the twin o.h.c. racing Salmson, or its touring counterpart, they are so similar. This compact long-stroke engine, with its hemispherical head and inclined valves, must be one of the outstanding small power units of the ‘twenties. The simplicity and the aptness of the design make it a classic, and there are ‘sufficient touches of originality to

indicate the mind of a designer of outstanding ability. It is difficult to see why the engine has m >t received more attention than it has, for it was certainly the first small twin o.h. camshaft engine to be made successfully in large numbers. The hemispherical combustion chamber combined with twin o.h.c. valve operation was first used by M. Henry in his 1912 G.P. Peugeot design, and later in more developed form in his 1920 G.P. Ballot. The twin-cam Salmson valve gear shows some

interesting similarities to M. Henry’s designs; and may be considered an intermediate stage. As on the Peugeot the camshaft runs through the valve-operating tappets, but the latter is in the form of a piston moving in a bush in the camshaft housing, thus removing from the valve stem any lateral loads; the bearing area of this arrangement is far greater than that of the earlier car, hut the p:stons are a considerable reciprocating. weight. By contrast the Lallot had small pistons interposed between the cams and the valve stems, which performed the same function with far less weight. This method is used today on most high.;performance twin-cam engines. The valves were inclined at 6o to each other.

The four cylinders were of 62 90 mm bore and stroke (1,og7 c.c.).. The block, of unusual design, consisted of top, middle, and bottom plates connected by the cylinders. The water jacket was either welded round the casting, (a method of construction favoured also by Bertarione on the 1922 G.P. Fiat), or On certain later models was removable and sealed by fastening down the detachable cylinder head on two rubber joints of circular section.

A vertical shaft at the front of the engine drove the camshaft by helical gears. At the base of the former a cross-shaft drove the twin magnetos mounted on extensions on either side of the crankcase.

The crankshaft was carried in two roller main bearings, with smaller subsidiary bearings at each end. Lubrication was by the well-tried pump and trough system, with two small crankshaftdriven pumps attending to the top part of the engine. The sump was an aluminium cast tray with finned base. A flat plate, with a gauze oil filter set into it, guarded against oil surge.

In contrast to the chassis, considerable engine development took place. From x922 to 1925 the engine was used by the Works team in unblown form. The compression ratio was raised from the modest 5.8 to i of the standard engine in the early days to almost 7,4 to t before the supercharger was fitted. The engine speed rose from 3,500 r.p.m. in 1921 until in 1927 4,500 and even 5,000 r.p.m. were being achieved. To keep pace with these developments the engine was extensively modified season by season. The lubrication, the cooling, the bearings, all received attention. At each stage of development production engines incorporating most of the works features were available to owners who were prepared to pay extra for them.

The early engines, using light alloy pistons and I-section connecting rods, were known as the ” Grand Prix ” series. They used pump and trough lubrication at first, but adopted a pressurefeed system, using a drilled crankshaft in later seasons. In 1923 the team was using engines with a large centre-roller bearing for the crankshaft, together with long fragile tubular connecting rods. A new lubrication system was used, in which numbers I and 4 big-end bearings were fed with oil from the pump directed into two concentric rings on the crankshaft and thence into the bearings. Numbers 2 and 3 bearings were fed with oil passed through an oilway in the centre main bearing, into rings on the crankshaft, and so into the bearings. The system, rather similar to that used by Bugatti on his racing engines, should correctly be described as ” semi forced feed lubrication.” Another modification was to the valve heads, to allow a longer dwell. Engines made to this specification were known as the ” Grand Prix Special ” series, and could be distinguished from the ” Grand l’rix ” series only by a small bolt head in the centre of the aluminium crankcase which served to locate the centre main bearing. In 1925 the engine was supercharged, and by the end of 1926 had become a very highly developed unit indeed. 8o b.h.p. at 4,000 r.p.m. was obtained by the Works, but it was claimed that

more power was available at higher speeds. It was on these engines that the detachable water jacket was employed. Magnesium alloy pistons were used and extremely careful balancing of all the engine components was necessary. The tubular connectingrods were inclined to ” bow ” slightly at very high engine speeds and the resulting collision with the wall of the narrow cylinder bore could be disastrous. A Cozette blower was the normal fitment, mounted low on the near side of the crankcase and driven at engine speed by a train of gears from the rear of the crankshaft. This method was claimed to prevent the transmission of crankshaft torsional vibrations to the blower, and to further this aim a flexible coupling was also used on the driving shaft just before the blower. A Roots blower could be obtained for the x928 models sold in this Country. Engines built to this specification were known as the ” San Sebastian ” series’ and could be identified by the 4-branch water-header pipe and the square ends to the cylinder block. Dual ignition was also standard. Typical of the attention to detail on these engines was the small half-bearing which steadied the centre of the camshafts. This had originally been a full bearing made from Dural and attached to the camshaft cover, but it was found that only the top half of the bearing surfac.: was being worn and so the lower half was dispensed with.

The braking system and suspension of these 50/15 models reached their highest point of development with the introduction of the ” San Sebastian ‘ model. Early chassis had the rear springs sliding in a trunnion above the axle; on the ” San Sebastian ” the axle casing was turned upside down so that the spring was under the axle and the chassis lowered by several inches. To balance this the front axle was a new wide-track unit of cast l-section with a deep drop from the king-pin to the spring pad. To this axle Perrot cable-operated to in. brakes, working in finned alloy drums with steel liners, were fitted. These brakes were more than the slender front springs could take, and considerable winding up occurred, exaggerating the servo effect of the large ” self-wrapping shoe in the drum. The result was that to achieve powerful snatch-free braking Was difficult, the technique being that of merely touching the brakes and letting the front springs do the rest. In contrast with the early models, Hartford type shockabsorbers were now employed. On some Brooklands cars quadruple units were fitted on each axle; the effect on the bumpy surface must have been to convert the car into a four-wheeled toboggan.

The production 10115 models were endowed with a great variety of bodies, ranging from sports two-seaters to lull sedanca de ville; the latter was at least catalogued, if not actually made. Unfortunately the chassis was over-bodied with most of these creations. Eventually larger engines were built, but by that time the bodies were even heavier.

But some light, attractive sporting bodies were also made. For the 1922 and 1923 seasons there was a handsome boat-decked 3-seater sports model. The third seat was a cockpit in the canoe tail, surrounded by small motor boat. handrails. Equipped with a fully disappearing hood and an adjustable windscreen, the car offered either or I-elliptic front springs. The price was £235 and the colours were yellow or white, with black wings. For saloon-car addicts there was an entirely delightful little ” Neo Gothic ” 2-seater fixed-head coupe. It had the fashionable vertical lines, with both square and oval windows. There was only one door, on the passenger’s side, and there was a luggage boot in the short tail. Finished in black the price for this little charmer, ” handsomely equipped,” was 250. guineas. An additional £14 was needed if a Ducellier self-starter was required. The price of the 9.5-h.p. model in open 2-seater form was £195.

The Olympia Show of 1923 saw the first announcement in this country of the Grand Prix 2-seater. These handsome little vehicles, based on the outline of the team racing cars, were to set the style of the small Continental sports car for some years to come.

They were available in fabric or metal covered form over an ash frame. Very narrow, with polished cowled radiators and long pointed tails, they did not have staggered seats as standard, but most owners used cushions at their backs to bring themselves forward, clear of their passengers The cockpit interiors resembled those of the light aircraft of the day, with their wellpadded sides and small ‘ee-screens. Weather equipment was usually no more than a diminutive hood fastened to two extending rods just in front of the windscreen. Equipment varied a great deal, but the instruments were usually confined to a rev.-counter, clock and ammeter. The more expensive models had fuller instrumentation. A brass-fronted Ducellier switch-box was mounted in the centre of the facia nand. An ingenious ignition control, connected to the magneto by a

series of cranks and cams, wts usually fitted just behind the steering. wheel, matched by the choke lever. At first long swept ‘vings were fitted but later cycle type were available, and neat semi-spatted

wings were fitted to the 192S Olympia ” San Sebastian ” model. Twin spare wheels were carried on either side of the scuttle, and there was a tiny door on the passenger’s side. The cheaper models

usually had Salmson centre-lock wire wheels, whilst the ” Special ” and ” San Sebastian ” used the Rudge Whitworth wheels. Tyre sizes increased from 710 .• 96 in 1923 up to the 715 I 15 ‘4 Ballots in 1927. The electrical equipment was either C.A.V. or Ducellier, and the price range L265 to 085. Richard Twelvetrees tested the G.P. model for MciToR SPORT in December 1926. This -car had been borrowed from George

Newman, who was an agent for Salmsons. Twelyetrees appeared to be under the impression that it was a practice hack, but it seems that the wily Newman had lent him a pretty-well-maintained example, for he had won his class in the Kent Sussex M.G. hill-climb in September with this same car. Twelvetrees had tested the earlier version of the G.P. the previous year in a muddy I mndon-to-Gloucester Trial and was prepared to be plastered with mud again on this test, but the swept wings had been dropped for semi-cycle type with valances joining them to the chassis, giving good protection. Conditions were very poor and testing was hampered by wheelspin. The steering was much better than on the previous model and only at very low speeds did heaviness make the absence of a differential felt. It was possible to change gear silently if touring changes were sufficient, but mute sporting efforts resulted in a certain amount of ” clashing ” and a firm hand was needed when dealing with the gearbox. The four-speed box offered ratios of I 31 to I, 71 to 1,41 to t and 31 to 1. Top speeds in each gear were 30 and 49, with 6s m.p.h. in third; the highest speed achieved was .84.12 m.p.h., which was held for halfa mile before slow-moving traffic necessitated slackening speed. Later 4,500 r.p.m. was held for over a mile in first gear t the example this performance set may help to explain who so few Salmsons have

survived! [We are not so brutal today It is obvious that these revolutions were not found in top, so we can only conjecture what speed this 1926 unblown eleven-hundred could achieve when wound up on a circuit like Brooklands. Finally, the car was tested up the 1-mile Succomb Hill from the Caterham Valley to UppCr Warlingham. There was a rough surface, two sharp corners, a maximum gradient of I in 3, sonse gulleys and a small stream running across, but a climb in 60 sec. was made despite wheelspin. The braking powers of the car received favourable comment. All-up weight was so cwt., equipped with hood, windscreen and electrical equipment.

In the ‘twenties racing and competition success carried real weight. Often this was the only means by which the motorist of the day could assess the relative merits of the many makes offered to him. This was especially true of the cyclecars, for here design had not yet coalesced and many bizarre little ears were to be seen. Salmsons therefore decided they would court publicity by competing in as many events as possible.

They were successful from the beginning. For six • years they won time and time again, first with the 105 cars, later with the ” San Sebastian ” blown models, and not until the advent of the exquisite Amilcar ” six ” were the Salmson “fours” dislodged. From the initial success in the 1921 Cyclecar G.P. they proceeded to their crowning achievements of the succeSsive elass victories, in the 1926 and 1927 ‘I’arga Florio races in the hands of Geri and Borzacchini. In this 1921 season three important victories were to set the style for the years to come. First there was the Swiss Six Days’ Trial held over an arduous Alpine route crossing the passes of the Grimsell and Furka. Much of the course was on tracks whick were often little better than the dry beds of streams. Lombard’s ” Type Sport ” four-push-rod car took first place after a run during which the only involuntary stops were caused by punctures. In July the G.P. des Cyclecars at Boulogne was dominated by the racingbodied G.N.-based twins. After averaging 49.8 m.p.h. for the 2130 miles the Salmson team carried away the Coupe Pickett. donated by the British G.N. exponent, who must have been a little irked by this result. Lombard was second and Bueno third. In September Lombard won the 1,100c.e, race at Le Mans with the prototype twin-cam car, the only Salmson entered. This was a neatly streamlined machine with a short rounded tail and a slightly sloping, flat-fronted radiator cowl. The engine was the 62 > 90 version, fitted with a double Zenith carburet ter, one magneto and a cone clutch. The weight was claimed to be only 6 cwt. unladen. The flared front wings were fitted to this model as well as the G.N. Salmsons, Lombard won at an average speed of 54.8 m.p.h., nearly tti minutes ahead of the second car, despite losing twenty minutes after the mechanic had failed to 1.11 the fuel tank fully. ” This false manoeuvre,” commented La .Vie Automobile drily, “enabled him to show that he had the Makings of a

• recordman de course.’ ” After his stop Lombard covered several laps at more than 60 m.p.h. average. lie was less successful later its the year when he brought this car to Brooklands for the J.C.C. 200 Mile Race. After leading the race he damaged two wheels at a pit-stop and had to be content with 2nd place behind Frazer-Nash its his G,N. This was no doubt a consolation to Mr. Pickett, Lombard gained his consolation by taking the too miles Class K record at 72.8 m.p.h. during the course of this outercircuit race. The first season had been a promising start and for 1922 the team returned to the arena well organised and with the resources of a large engineering organisation behind them. A policy was laid down and was to be followed for the whole period the works rated seriously. For the most important races the works team would enter with their own drivers and cars. For secondary events “

Service de Course ” cars would be lent to an agent or driver of acceptable ability, often backed by works mechanics, part of the expenses falling on the driver. The remaining events deemed worthy of interest by the works would be supported by a driver using Isis own car with works advice and, occasionally, assistance.

In February, before the season proper opened, the French Light Car Trial was held over a difficult 152-mile course in the Forest of Marly. Conditions were miserable and it was not easy to maintain the required average speed of 24 m.p.h. Most Of the entries were of cars with special lightweight bodies and only the sketchiest of wings, the result being that their drivers rapidly became plastered with mud and finished the course in acute discomfort. Salmson fortunately had entered fairly standard cars with adequate mudguarcling, so their drivers fluent?. I lamel and Lombard had a comparatively comfortable ride, finishing in the first three places. The new team racing cars were fitted with neatly streamlined high slim bodies. The driver and mechanic sat well down in the cockpits, with only their heads in the slipstream. Whilst the latter squeezed himself in as best he could, the driver had adequate space. These bodies were really light and were almost of stressedskin construction, for the only formers were two hoops, one its front and one behind the cockpit. Almost circular in section, they relied on their curved surfaces to give them stiffness. Fulllength undertrays were fitted. Inside the cockpits the cars were very spartan t the seat backs were minimal and supplemented by webbing Straps. The weight of the car was 61 cwt. standing on its 700 • 85 tyres. These bodies offered excellent protection for driver and mechanic. W. F. Bradley commented favourably on the fact that the occupants were completely concealed within the barrel-shaped body should the car overturn. These cars were to be raced regularly until the end of the 1925 season,when they were made obsolescent by the introduction of the blown” San Sebastian model. It was one of these cars in single-seater form which was to record the first 100 rts.p.h. its the hio.o-c.c. class. They had a

habit of finishing in team formation, and on more than a dozen occasions took the first three places, occasionally heading the general classification. Salmsons won hundreds of races and it is sufficient to mention only some of the more important or interesting events.

In the 1922 season the superb Robert 13enttist, then at the start of his career, led the team to the first two places of the 1,10o-c.c.

class of the G.P. des Voiturettes at Le Mans at 61.3 m.p.h. average for the 264 miles. Dcsvaux was second. Benoist also won the 1,too-c.c. class of the J.C.C. 200 Mile Race at the respectable average of 81.88 m.p.h., in an unblown two-seater. Desvaux was once More second. The winner was said to have the engine from the single-seater which had taken the A.C.U. Class K flying kilometre record at 91.04 m.p.h., and was lapping-at over go m.p.h. during one stage in the race. Later, at a sprint meeting in the Bois de Boulogne, Lombard was timed in the ” razor blade ” single-seater, with wheel discs, at 94.9 m.p.h., for the kilometre, Benoist managed 82.8 m.p.h. in the two-seater.

1923 was the year of consolidation. In their class they were as dominant as the ” Invincible” Talbot Darracqs were in theirs. Benoist continued his winning ways in the 0.13. de Boulogne races, taking the 1,100-c.c. class at 66.3 m.p.h., and finishing second in the general classification. Bueno was second but the third car in the team retired, so the V-twin Frazer Nash team was able to retrieve the Pickett Cup from Salrnsons. It was Benoist again at Le Mans, where he won the 1,loo-c.c. class of the 0.1′. .des Voiturettes at 65.5 m.p.h. for the 247 miles, with a record lap of 70.4 m.p.h. In this race a 750-C.c. Salmson driven by Lombard won the smallest class at 62.5 m.p.h. for 182 miles. The sports models won both the 24-hour races held. At Le Mans, driven by Desvaux and Casse at 43.8 m.p.h. (12 h in General Classification) and followed by Bueno and Benoist (15th in General Classification), and at the 1301 d’Or where Desvaux and Benoist staged a dead heat, covering 1,104 miles at 46 m.p.h., with Casse third, after driving single-handed for the 24 hours. The cars ran stripped of wings. An interesting event was the fuel consumption G.P. de Tourisme. The two-seater light cars had to :average 47.1 m.p.g. at 5o m.p.h., for 184 miles. Desvaux’s Salmson finished third behind INVO Mathis after Benoist had retired with radiator troubles after leading most of the race.

By 1924 the Salmon was firmly established as a leading light car and the number of soon-to-be famous drivers who used them in their apprenticeship is an impressive list. Salmsons announced their retirement from racing in mid-season but still contrived to win a few races, including the J.C.C. 200 Mile Race with 0. WilsonJones driving, at 85.7 m.p.h. average. The new blown engine was now being used in the team cars in some races.

The 1925 San Sebastian G. P. races saw the debut of new racing models and Casse won the i,too-c.e. class with this car, which was accordingly titled the ” San Sebastian.” At three sprint meetings too m.p.h. was exceeded for the first time by an ” elevenhundred ” (‘foul-Nancy at tot m.p.h.. Verneuil at lo2.25 m.p.h., and by Casse in the single-seater at Arpajon Ot 109.2 m.p.h. ti.71. tue kilometre). So Salmson were the first-ever too-m.p.h. cars in the class and take their place between the too-m.p.h. A.C. and the 1.00-m.p.h. M.0. The racing-car events of 1926 saw the team still successfully using the blown ” fburs,” hut their time was nearly up, for the 0.6

Amilcar was obviously going to be a challenge. Nonetheless, Casse and Rousseau began the series of runs in the 24-hour race at Le Mans, which ensures the marque’s place in the record books even today, when they finished ninth in the classification and the first eleven-hundred on distance, qualifying for tile Coupe Biennial. The works ” San Sebastian ” team dominated the G.P. de Boulogne where Boudon won the class. Salmsons were prominent at Brooklands that year. I lazlehurst

won the J.C.C. Production Car Race outright, at 62.9 m.p.h., in his 0.1′. car, after his ” San Sebastian ” had been refused an entry because too few had been built. In the Essex M.C. Six Hours’ Sports-Car Race Newman won the r,roo-c.c. class at 50.1 m.p.h., and Goutte had brought the single-seater works car over from France and recorded a lap at 114.9 m.p.h. during a short Brooklands race in April, a performance which has since become, quite rightly, famous.

Far away in Sicily I3orzachinni finished first of the elevenhundreds at 38.38 m.p.h.; this was in its way equally meritorious’ as a perforniance. Despite this the new Amilcar finished in the first three places in the J.C.C. 200 Mile Race.

In 1927 the major success was at Le Mans where de Victor and Hasley confirmed the Coupe Biennial (Index of Performance) prize at 52.82 m.p.h., followed by Casse and Rousseau. In Sicily Geri brought his Salmon home first eleven-hundred again. In the 1928 Le Mans 24-hour race Casse and Rousseau won the Coute Biennia/ once again, at 57.177 m.p.h. Private owners continued to win many races but Le Mans 1928 was really the last major event won by the team. Amilcar produced the” sixes ” in answer to the Salms,m ” blurs” and in pursuit Of even greater piston area Salmson embarked on the design of a straight-eight of 49.9 70 mm. early in r927. It was intended to run up to 8,000 r.p.m., and developed too b.h.p. at 5,800 r.p.m. It has been said that the engine had desmodromic valve operation, but this is probably due to misreading the rather vague contemporary technical reports. Some of the existing cars were modified to accept the new power units but the teething troubles were prolonged. They had a few successes in hill-climbs and speed trials and one car came over to Brooklands and was unofficially timed at 137 m.p.h., but they were not generally satisfactory. It was entered for three raves at the Easter 1928 B.A.R.C. Meeting by A. Bouvier, surely the smallest car to be listed among ” Lightning” Handicap entrants, but it failed to appear. Later that season Newman’s 4-cylinder car lapped at 103.76 m.p.h.–ED.] A new version of the car appeared in 5929 and won the Gaillon Hill-Climb in the hands of Yves Giraud Cabantous, who afterwards became famous in LagoTalbots in the ‘forties. ‘[‘his was a very low offset single-seater with the radiator slung between the dumb-irons. The chassis was of double-drop type and there was a tall headrest for the driver, but by this time the factory was concentrating on more sober models and did not persevere with the type, although a similar car raced at Montlhery in the ‘thirties driven by R. Girod. It is difficult to obtain much information on these ears and also on the four-cylinder cars entered fOr the 1928 Le Mans race, which were described as having roller-hearing big-ends, a feature they shared with Bartlett’s Salmson, a successful Brooklands Mountain Circuit car, and one which made these unlike any other Salmsons. [In 1932 Bartlett twice broke the Mountain Circuit lap record, and he did so again in 1933, leaving it at 69.95 m.p.h., until he lost it to Straight’s M.0.—EDI That concludesthe story of the vintage Grand l’rix Salmsons. Very few are left now. Our French Salmson Owners Club lists less than thirty ears and a large proportion of these are no longer original, their fabric bodies having long given best to our damp

climate. Nonetheless, it is a proud story and one that certainly justified their British concessionaires’ slogan ” Salmson —The car that wins.”