A Discourse on the World’s Best Car at the Time of the Armistice
A lengthy and at times heated discussion took place in the correspondence columns of The Autocar on the subject of which was the best motor car money could buy in the formative and exciting period of innovation and development after the end of the Kaiser War— roughly from the Armistice of 1918 until the Industry got more or less into its stride by 1922. This correspondence ran for not far short of a year and as so much interest was taken in the matter of the premier luxury-car of the immediate post-war years by contemporary commentators, it must surely be a subject which present-time motoring historians cannot ignore ?
In the course of the aforesaid correspondence the late J. G. Parry Thomas, Welsh designer/racing driver, proclaimed that the only car he recognised as a sparring partner for his Leyland Eight was the Rolls-Royce. That was the opinion of the then-Chief Engineer of Leyland Motors Limited. That being the case, it seems just to compare this post-war design with the other top motor cars of the day and see how it rates.
Which cars shall we line up beside the Parry Thomas creation ? I suggest that they should be ranked as follows:
Hispano-Suiza 37.2 h.p. .. Designed by Marc Birkigt.
Isotta-Fraschini 36.4 h.p. .. Giustino Gattaneo
Lanchester 40 .. George Lanchester
Napier 40/50 h.p. .. A. J. Rowledge
Rolls-Royce 40/50 h.p. .. Sir Henry Royce
It may be argued that some significant luxury products have been omitted. The Daimler, for example, used by the Royal Family. But the 45-h.p. Daimler of the period under review was virtually a pre-war design, with a sleeve-valve engine of great complexity, and a car intended as a comfortable carriage with no pretence to performance. Of the six cars listed, all were new since the war except for the well-tried 40/50 Rolls-Royce Silver Ghost, which had already earned for itself, or, if you prefer, had acquired for itself, the proud title of “The Best Car in the World”. It is thus impossible to exclude the car made in Derby. But I would ask my readers not to be prejudiced by its advertised perfection—although no doubt most of them are!
There are other cars you might have expected me to include in this communal sorting-out. The Farman, the Sheffield-Simplex, the Delaunay-Belleville, the Ensign and so on. But none of these quite came into the category of widely-discussed, new, exciting cars, combining dignity with some degree of performance. I am aware that I might have put in the very excellent 24-h.p. Delage, and maybe the 37.2-h.p. Hotchkiss. And what of Louis Coatalen’s Sunbeams ? The fact is that, while there were at this period a great many good cars available, we are engaged in trying to decide how the Leyland Eight, the first British straight-eight production car, rated against top-bracket contemporaries. While to list these on price alone would be dangerous ground to tread; I think this arbitrary criterion must be to some degree the yardstick to use in keeping the makes listed to manageable proportions.
When the doors of the 1920 London Motor Show opened, the first at which the Leyland Eight had its public debut, immediately being dubbed “The Lion of Olympia”, the Sunbeam was still a staid side-valve Edwardian-style chassis and the others were not mentioned freely when discussions on the topic took place. The fact is that Armstrong Siddelcy, Crossley, Fiat, Lancia, Renault, Spyker, Straker-Squire, Vauxhall, Voisin, Wolseley and others made good cars and had certainly not neglected the carriage-trade. But we are here concerned with the best. Also, please remember the period with which this discourse deals. The fascinating era just after the First World War, when war profiteers and some others had much money to spend on new cars and had ambitious fresh designs, showing aero-engine influences in their under-bonnet specifications, from which to make a choice, was one in which some later top models had not so far emerged, or had yet to attain full eminence. This was true, in insular Britain at any rate, of the American cars, of the Bentley, which had scarcely got going in sporting 3-litre guise and was only later to flower as the magnificent Big Six and splendid 8-litre, and of the Talbot-Darracq, then in no way the benefactor of Georges Roesch’s initiative and the vee-eight cylinder power unit of which was regarded as too unusual to be instantly accepted in this country.
In later times, too, Daimler came out with the modernised “Double-Six” of revered memory, in Belgium Sylvain de Jong announced the 8/40 Minerva and the Excelsior had become known here, Mercedes-Benz followed their immortal 36/220 and 38/250 sports cars with Nazi-ware of up to “Grosser” size and substantiality, Ettore had been tempted from the circuits to evolve the Bugatti Royale, and Delage had progressed to the D8, Isotta-Fraschini to the 8A and 8B. But by then Rolls-Royce had come round to overhead valves and four-wheel-brakes, Hispano-Suiza had gone to 45 h.p. and eventually used twelve cylinders, and if Leyland and Napier had forsaken their early promise in favour of lorries and aviation-propulsion, Lanchester gave us the impeccable straight-eight 30 before being commercially engulfed. And as the multi-cylinder engine in both in-line and compacted vee formation came to be accepted, autos like Cadillac, Lincoln, Packard and Duesenberg entered the top-league stakes. But this was in the future, and although it caused the “Best-Car” controversy to break out again in the letters-pages of The Autocar, many of the Utopian layouts then discussed owed little or nothing to the Armistice-age designs.
So we can, I hope, proceed, within the limits of the listed makes, those built, while the Empire was still licking its war wounds, in Bois-Colombes, the via Monterossa, Birmingham, Leyland in Lancashire, Acton in London, and Derby.
They are listed above as they presented themselves to rich customers in October, 1920.
Those are the runners and there will be many, if they have got this far, who will say I am wasting their time, because there is no possible doubt as to which was the World’s top car in 1920, no possible doubt whatsoever —the Rolls-Royce Silver Ghost. I would agree that it was certainly the most popular of the great motor cars of that time. Its reputation had been sealed thirteen years before. It had a tremendous war record. Many wealthy motorists bought it because of this repute. They often made do with pre-1914 Ghosts, or bid to higher than the market value of a new one even for examples that had seen service on the Western Front, while waiting patiently for their 1920 chassis to arrive. But there is a difference between quantity and technical superiority. The Rolls-Royce was available and very much a known design in 1920, when others, certainly Leyland and Napier, would not be so for another couple of years. But how antiquated the Ghost was! It dated bad; in the metal a matter of 14 years. It had a side-valve power unit which a French paper had likened to a Christmas tree (or was it the spreading chest-nut tree ?), so untidy were its externals. Someone else had called it a triumph of workmanship over design. It was arguable that in quality and constructional execution the cars from the Derby factory were supreme. But were such standards beyond duplication by such great factories as those of Lanchester, Napier and Hispano-Suiza ? We may allow that Isotta-Fraschini was rather an unknown quantity and that lorries were perhaps not to be likened to quality cars—but that is begging the question of whether Parry Thomas would allow the rough and ready of commercial vehicle manufacture to undermine his masterpiece, .a masterpiece conceived from an electrical and aero-engine background by one rumoured to be much more engineer than motorist at that time. Apart,of course, from the fact that those who visited pre-war Commercial Motor Shows used to remark that the engineering there to be seen made private cars look unhappily crude. . . .! If such is sufficient reason to merit continuing, let us look at each design, stage by stage.
1. Point is given to this daring suggestion by use fact that after 1918 Lancheater made their own road springs, with greater precision than the professional spring makers, and machined the entire front axle, by the manner in which.Hispano-Suiza machined the crankshaft of the 37.2.h.p. engine from a solid billet of the finest steel procurable, of which 671 lb. was machined from the solid billet to produce a circular-web shaft weighing 99 lb. and finished to a jeweller’s degree of accuracy (there was also Hispano-Suiza’s method of rendering the upper half of its light-alloy engine corrosion-proof by a patented process of enamelling under pressure), and the aeroplane-engine methods of construction employed by Napier in whose factory the 40/50 was made under the supervision of established engineers and designed by an ex-Rolls-Royce mom.
It is obvious that whether they knew the saying or not, that “there is no substitute for litres”, all the 1920s designers of luxury motor cars employed plenty of capacity. The six power units we are studying ranged from 5.9-litres to 7.4-litres. Not unnaturally, RollsRoyce, as befitted a virtually ‘Edwardian concept, had the most swept volume. But Thomas, using a very modern valve gear, went to almost the same size of engine, so Can be said to have made sure of both baking his cake and eating it! Mark you, by having eight instead of only six cylinders he was justified, in terms of piston area and reasonably compact combustion chambers. Indeed, it is surely significant that apart from Isotta-Fraschint’s much smaller eight, Thomas had the smallest cylinder bore, which he retained when he decided to increase the capacity of his engine. And I need hardly add that at a period when, although each of these six top luxury-automobile designers felt that four cylinders were too few for their purpose, only the broad-shouldered Welshman and an Italian ventured to have eight, Thomas being the first British designer to do so for a catalogue car.
It is interesting that although we are told that the average customer did not trust overhead valves in this early-vintage era, only Rolls-Royce had side-by-side valves and was to discard them within five years. Aeroplane engine rather than motor-racing precedents are said to have caused the creators of our five new luxury chassis to opt for upstairs valves. So be it! Note that all except Isotta-Fraschini also put the camshaft above the head.
I now ask you to observe how such advanced valve gear was arranged. The drawback to using o.h.c. valve gear for a gentleman’s carriage was noise. The chores of’ tappet adjustment and decarbonising which this layout made difficult was the chauffeur’s worry. If it defeated him, he could perhaps persuade his affluent master to let the factory do the job and charge for it. But noise! Especially when the Ghost truly lived up to its name. Yet Marc Birkigt calmly endowed his new motor-car engine with the valve gear he had perfected for V8 aero-engine use. A vertical shaft at the front of the engine drove the o.h. camshaft by bevel gears. True, the rest of the gear Was brilliant, the cams operating direct onto adjustable discs on the valve stems, neat, efficient and simple, and admirably suited to engines required to turn at much higher speeds than a liveried chauffeur would use. Indeed I am already inclined to question whether Birkigi had performance rather than luxury in mind when he planned his great post-war car; if this seems so, we might thus early eliminate the Hispano-Suiza from the. Context of the present study. Only its Worlddemand by those more interested in the latter than the former, causes me to retain it!
George Lanchester used skew, or worm, gears in his o.,h.c. valve gear for the post-war Lanchester 40, a form of drive with which he was amply conversant. It was quieter presumably than a bevel drive for the front-located vertical shaft he likewise employed but heavier side loads had to be allowed for. Lanchester had a rather more conventional head, with mildly inclined valves, than Birkigt, rockers being imposed between the cams and their stems. A. J. Rowledge, who had departed from Rolls-Royce Ltd. to give Napier a fine new post-Armistice car, went in for very similar valve gear, with his vertical shaft also driven by worm gears, but with his valves upright instead of being inclined. Isona-Fraschini relied on push-rod o.h.v.; the Rolls-Royce had an L-head side-valve. The rockers used to operate these o.h. valves may have been of ingenious shape and the Napier’s had inset rollers for the cams to bear on, but they were nevertheless of conventional conception.
Regard, however, the Leyland Eight’s valve gear. Thomas had no intention of allowing his overhead camshaft drive to cause cacophony which would distress the ears of the Leyland’s occupants. He drove it with triple eccentrics, set at 120 deg., and coupling rods, from the rear of the engine—a silent drive which must be accepted as superior to all the others, for did not W. O. Bentley adopt a similar system for the 6.1/2-litre and 8-litre Bentley engines ? Although rumour suggests that he was not quite so adept at preventing expansion and contraction affecting the drive, as Thomas had been! Not only that, but Thomas used rigid girder-type valve rockers, with, beneath them, leaf valve springs, one for each pair of lightweight tulip valves. At this time it was common knowledge that coilsprings could lose their temper and cause valve flutter, or even break, particularly if subjected to the heat of an overhead-valve engine. The motorcycle fraternity made use in some engines of hairpin springs, for this reason. But as long ago as 1918 Parry Thomas had decided on a substitute for the conventional valve spring. I suggest that the valve gear of his Leyland Eight was definitely an advance on those o.h.c. designs of Hispano-Suiza, Lanchester and Napier.
Of other aspects of engine design, all these grande luxe cars had most of the expected refinements, and Rolls-Royce, of course, at that date made their own carburetter and electrical equipment. It is interesting that whereas Thomas was content to use Delco coil ignition, the others often went to the weight and expense of dual or duplicated ignition systems. Thus the Ghost had two sets of plugs fired by coil and magneto (the latter a Watford, not made by R-R), as had the Napier, and the Lanchester had magneto ignition with a coil as a standby. Like Thomas, Birkigt was happy with Delco coil ignition but he used duplicated distributors firing twelve plugs. As a safeguard against those unaccustomed to battery ignition leaving it switched-on after parking the car, Napier contrived a sort of hot-wire cut-out for their coil circuit but Parry Thomas took an engineer’s approach to this minor problem by having a suction cut-off for his ignition circuit, so that if the engine was not turning the battery was isolated.
All these engines except the Hispano-Suiza seem to have had fairly normal forged crankshafts, with mostly seven plain main bearings, although Lanchester used what amounted to nine, whereas Thomas was conservative, having only six for his in-line eight. Perhaps he was anxious to cut down frictional losses.
At this period there was some concern about big engines, operating in a wide temperature belt, boiling over or running too cool. The Ghost had been given a radiator thermostat by 1921, and other solutions were radiator shutters, operated either by hand or heat, and going outside our present context, provision on the Farman for the driver to engage or disengage the fan as the thermometer indicated. Napier found they had to fit a thermostat, accommodated behind the radiator in a rather ugly fashion, some time after the 40/50 had been announced. As you may by now have come to anticipate, Thomas used a neat solution on the Leyland, water flow being controlled by a thermostat and the fan being put out of action at small throttle openings, by suction control.
Most of these engines appeared originally with single carburetters, Napier painstakingly developing the SU to their requirements. But Thomas made provision for twin carburetters, and the Isottas seem to have had them from the start. Constructionally, the classic new methods were those of Hispano-Suiza and Napier, with aluminium for cylinder block and head, and inserted steel or iron cylinder liners. These were the aircraftpattern power units which were generally distrusted by conservative British buyers in 1920, especially as the French engine had separate cylinder barrels with integral heads. Isotta-Fraschini also had a light-alloy block. The others were more conservative—perhaps costly weight-saving was not thought necessary in cars destined to carry heavy bodywork —but aluminium pistons were now universal, pioneered before the war, we are told, by W. O. Bentley on his DFPs. But the shadow of Edwardianism fell on some of these great luxury automobiles, as is evident by the triple cylinder blocks of the Ghost and the paired blocks of the Lanchester, and the non-detachable heads of three of the engines. Tubular con.-rods were used by Leyland, Napier, Isotta-Fraschini and Hispano-Suiza.
With such cars as I have been describing, absolute power output was not the sole criterion, smoothness, silence and low-speed torque being also of great importance. But nevertheless we must take stock of the power outputs claimed for these engines, before passing on to the chassis. They can be stated as:
Hispano-Suiza: 135 b.h.p. at 3,000 r.p.m.
Isotta-Fraschini: 75 to 80 b.h.p. at 2,200 r.p.m.
Lanchester: 89 b.h.p. at 2,000 r.p.m.
Leyland (single-carburetter): 115 b.h.p. at 2,500 r.p.m.
Napier: 82 b.h.p. at 2,000 r.p.m.
There was some variety even in the matter of getting power from the engines just described to the back wheels, over which future dukes and duchesses might well ride. Rolls-Royce had a cone clutch and a none-too-easy-to-operate 4-speed gearbox. The Isotta-Fraschini, like the Hispano-Suiza, had only three forward speeds, which was something of a handicap on rolling English roads (or so Englishmen were apt to persuade themselves if they had just paid handsomely for a Ghost), both these cars being intended primarily for devouring long, straight highways. Both had the new-fangled unit construction. But whereas Birkigt had a cone clutch (having made this component work smoothly from Alfonso days), the Isotta had a multi-plate clutch. The Ilispano-Suiza change was delightful, providing the lever was whipped through the gate. George Lanchester gave his customers a 3-speed epicyclic box, but with a normal gear-lever for stirring it. It was a whine-er if the engine was allowed to idle in neutral. The Napier was conventional, with a plate clutch, except that the gear-lever was in the centre, whereas it was the fashion for chauffeur-driven cars to have r.h. levers (so that the paid-man should not fumble about near the knees of the housemaid or, worse, those of the daughter-of-the-house ?). Parry Thomas wanted nothing out of the ordinary here, employing a plate clutch and a 4-speed box, with a r.h. lever. If Lanchester used oil under pressure for his gears, perhaps important with sun-and-planet wheels, Thomas arranged a gravity supply to his gearbox. He also arranged for the gearlever to start the engine when it was put into the appropriate slot. This obviated the engine being started when the Leyland was in gear. But the real reason for this was probably that, as the starter motor drove via the gears in order to provide extra torque for turning a big, gummy engine, by having the starter-contacts close at hand Thomas saved on cable, an economy aimed not so much at saving cost as voltage-drop.
It can be said that the frames of all these cars, with the exception of that of the Leyland Eight, were Edwardian in conception. A staunch supporter of George Lanchester has admitted this to be so with the post-war Forty, even if tubular cross-members stiffened it to some extent. The Napier was given extra-deep side-members but it was not otherwise particularly stiff. Both these cars, and the Ghost, had the much-favoured cantilever back springs, giving the greatest possible spring-base, the Lanchester’s spring leaves precision-made in the factory and its back axle, retaining a worm drive. Birkigt and Cattaneo had half-elliptics all round, as befitted fast cars intended for considerable development.
We now come to the most outstanding of Parry Thomas’ design features. He gave the Leyland Eight a very advanced system of suspension, using a supple springing medium quite unusual in 1919, which his very rigid frame with enormously deep side-members made possible. (He had patented this Thomas suspension, so when he left Leyland Ltd., had the Company gone on with car production they would presumably have been very much in the Welshman’s pocket.) So rigid was the Leyland chassis that it was provided with extensive lightening holes. The front axle was Sprung on unshackled half-elliptic springs but stout radius arms located the axle and acted as anti-roll torsionbars. At the rear, quarter-elliptic springs functioned in conjunction with true transverse torsion-bars. Thomas thus not only pioneered torsion-bar suspension but used it to achieve true anti-roll properties. it is possible to read of other steps being taken to combat, to a minor extent, roll in cars with long cantilever back springs and top-heavy carriage-work. But none was so effective as Thomas’ remarkable layout. It stood him in pod stead when he built racing cars to lap Brooklands at well over 120 m.p.h. Most of the big cars had torque-tube transmission (Rens-Royce retained the Hotchkiss-drive), but Thomas showed more ingenuity here. He mounted his differential on the end of the torque-tube and was thus able to have the drive-shafts splayed in a shallow vee, an attempt to maintain negative camber of the back wheels on the convex roads of those times, when flat arterial surfaces were few and far between. (This is said to have served Thomas well when he ran his cars at Brooklands, as by, in effect, turning the axle over he made the wheel camber suitable for the Track’s concave bankings!)
Although it has been expedient to cover chassis frames and -suspension under one heading, these being, in Thomas’ case at any rate, complimentary, brakes deserve separate study. At the period under review the English would have little to do with front-wheelbrakes, fearing the terrible consequences of a wheel locking-on while cornering, although some years prior to the war Argyll and Crossley had shown this to be a fallacy. (All credit to Sir Herbert Austin in using them on his little Seven by 1923.) So the employment of four-wheel-brakes, and applied by mechanical-servo at that, on the Isotta and Hispano cars, placed them right ahead in respect of safe retardation. The others of our elite half-dozen relied on old-fashioned rear brakes and both Lanchester and Napier even had a foot-applied transmission brake.
Much has been made of how Royce had to go to Hispano-Suiza for his servo and then could not at first get it right (those who would know more of this should read Mr. Robotham’s excellent book). The fact remains that Isotta-Fraschini had fitted 4WB before the war, and so may have known even more about them at this stage than Birkigt did, but the gearbox-servo was being made good use of by Renault, Sunbeam and others by the early 1920s. Thomas had not-come up to such standards but he did have a vacuumservo to apply his foot-brake, and is rumoured to have sold the patent for it to Clayton-Dewandre, this system becoming extremely popular. Later five of the cars under review were to adopt all-wheel braking, Lanchester even having a power-servo for non-hydraulic brakes, which in a sense pre-dated Citroen and Rolls-Royce with the Silver Shadow in this respect. I have seen a faded photograph which suggests that Thomas, too, may have put a front-braked axle On a Leyland Eight, but of this I cannot be sure.
Great cars should have their distinguishing marks (although there were those devoid of badges), and these we are dissecting were no exception. R-R had the famous “Spirit of Ecstasy” radiator-cap mascot, Ilispano-Suiza the equally famous flying stork “La Cigogne Volante”. The Lanchester had the practical feature of a water-level glass in the top of the radiator, Napier the “water-tower” radiator filler cap. The Leyland had for a rime square-shape headlamps and conical hub-caps for its Michelin disc wheels, but luckily not on every model. However, the greatest designers can err; HM King George’s comment to George Lanchester when viewing the ornate interior of the first Lanchester 40 was: “More suited to a prostitute than to a Prince, don’t you think, Mr. Lanchester ?”
Further Leyland Eight Niceties
Apart from some of the points already dealt with, Parry Thomas had a four-gallon gravity-f:d reserve petrol supply, an automatic chassis lubrication system actuated by gravity and also by the movement of the rear road springs (which Rolls-Royce in America, with a different system, did not fit until 1926, and Derby not until 1929). He insulated his gearbox by mounting it on leather rings. And whereas the other o.h.c. engines we are concerned with had vertical or very slightly inclined valves as in the Lanchester, Thomas had true hemispherical combustion chambers, and it is said that he obtained a slight desmodromic effect from the action of his leaf-valve springs, for his light tulip-shaped valves. The little-ends were lubricated via the tubular con.rods and instead of carrying the engine oil in a conventional sump, Thomas used a reservoir cast on the o/s of the engine, to which the flywheel took the lubricant, for distribution by high-pressure pump. The engine compartment was kept free of the steering column by using two steering boxes, which also permitted a steep rake for the wheel. Standard equipment included twin-horns and dipping headlamps. There were only two oilers on the whole chassis. Incidentally, by the end of 1921 power output was quoted as 60 b.h.p. at 1,000 r.p.m., 100 b.h.p. at 1,800 r.p.m., with a maximum of approximately 145 b.h.p. The chassis was to be available in three different wheelbase lengths.
So much for the design and construction of these fine motor cars. In my opinion (and yours ?) the Leyland Eight was the best car of its time, at all events on drawing-board form. In the next instalment it will be my task to investigate how they appealed to those who took them out on the road, and how these six cars performed. For the real proof of a car is in the going thereof. . . .—W.B.