Since, fortunately for the Editor, nobody is likely to regard anything appearing in “Sideslips” as a very clear reflection of Editorial opinion, I am, I conceive, one of the few writers on motoring subjects who is in a position to say that he found the recent Motor Show at Earls Court dull. I do not expect that any of the New Look stylists will mind much, anyhow; they will merely write “Baladeur” off as an old fogey who hasn’t noticed anything that has happened in the last twenty years or so. As a matter of fact they will be wrong. I have very definite views on this subject, and I consider that Saoutchik is the only coachbuilder, among those whose créations I have been privileged to inspect, who has made anything at all of this modern idiom, however much the pundits may laud the Italians. I believe that if Saoutchik goes on as he is going, he may yet achieve something, entirely in the modern manner, as good as anything that, say, Kellner produced in the ‘twenties. But that is entirely beside the point.
What I set out with the intention of saying was that one of the few exhibits at Earls Court which I cared to look twice at was something without any coachwork at all, namely the E.R.A.-Javelin chassis. This, I thought, may turn out to be one of the outstanding high-performance cars of the present epoch, and if it does, it will be particularly interesting because it has a horizontal engine. In this respect it is in increasingly impressive company. At the Paris Salon, there was not only the new flat-four hatchkiss-Grégoire and the new flat-six Matthis, but the flat-twin Citroën which has been seen before and the flat-twin Dyna-Panhard which is already well-established. Most significant of all, the flat-twin Dyna-Panhard . . .
In the early days of the motor-car, the appeal of the horizontal engine was a siren song which lured more designers, if not to their destruction, at least to their grave disadvantage, than I suppose, any other temptation to do what appeared to be eminently reasonable. I do not know quite why, when Carl Benz set out in 1885 to make his first “oil spirit motor tricycle carriage” at Mannheim, he decided upon a horizontal engine, because he immediately threw away what appears to be its major advantage by giving it a vertical crankshaft and a horizontal flywheel. As a result, the latter rotated in a plane at right-angles to the road wheels (in the azimuthal plane, is, I believe, a smarter way of saying the same thing) so that Benz had to introduce a bevel gear into his system in order to sort things out in this respect. However, he quickly saw the error of this, and as soon as he had given his horizontal cylinder a horizontal crankshaft, he had everything revolving in the same plane, and all bevel gears could be dispensed with. Perhaps, unfortunately, he thereupon succumbed to the temptation to use belts for the drive from the crankshaft to the countershaft, but that was not fundamentally the fault of the horizontal engine, and on the whole his solution of the main problem must have seemed to be the sensible one.
His contemporary, Gottlieb Daimler, preferred what was in effect an almost by vertical engine, the two cylinders being in fact arranged in a narrow V, but when, with the assistance of Maybach, he fitted it into the early Cannstalt Daimler chassis, he arranged it with its horizontal crankshaft set across the car, so that the effect was much the same as in the Benz, and the drive was taken to the countershaft by means of belts, without the interposition of any bevel gears.
No sooner, however, had the Frenchman Emile Levassor been presented with the Daimler engine as the power unit for a motor-car, than he proceeded to behave with it in what was apparently about the most perverse manner possible. The Germans, once they had got over their original predilection for tricycles, had proceeded to build vehicles on the lines of a four-wheeled dog-cart; and as they did not particularly want to carry dogs in them, they decided, as it must have seemed very sensibly, to use the box at the back underneath the seat, which was provided for this purpose, for the bestowal of the engine. There it was out of the way (all too much out of the way, as a matter of fact, when it went wrong) and at the same time handy for its work of driving the back wheels. Levassor, however, although he agreed that the back wheels were the right ones to be driven, placed his engine right at the front of the car, where it was as far away from them as possible. He seems, however, thus early to have argued that whereas a dog-cart has a horse at the front to keep therfore wheels on the ground and pull them round for steering, a homeless carriage has not, and, if the steering wheels are to have adequate adhesion, they must have a pretty heavy weight mounted pretty close to them. The handiest thing he could find for this purpose was the engine.
But having thus, for reasons which seemed to him sufficient, located it at what the Germans regarded as the wrong end of the car, he proceeded to mount it the wrong way round, that is to say, with the axis of its crankshaft parallel to the longitudinal axis of the car. Having then proceeded to produce the line of the crankshaft in the form of a clutch-shaft carrying the gearwheels of his brutal sliding-pinion change-speed gear, he was faced with the necessity of introducing a bevel gear in order to translate its motion into that of a countershaft across the car, carrying on its ends the sprockets of the final driving chains. And having committed all these apparent follies, he had produced what is in essence the motor-car of to-day.
Or course at the time very few people saw eye-to-eye with him. He sold a number of his Daimler motors to Peugeot, but Peugeot, while taking Levassor’s word for it about the longitudinal crankshaft arrangement and the sliding-pinion change-speed gear, did at least put the engine at the back, whence it drove forward to the gearbox and back again by chains to the back wheels. At the time of the Paris-Rouen trials in 1894, therefore, and of the first motor-race from Paris to Bordeaux and back in 1895, Levassor was alone in his glory, for besides the Panhards and Peugeots, the only petrol cars which figured at all prominently in these competitions were the French editions of the Benz, such as the Roger and the Parisienne. Moreover, while these were already outclassed, there was as yet nothing to choose between Panhard and Peugeot; in Paris-Rouen the first prize was divided between them, and in Paris-Bordeaux-Paris, while Levassor made the fastest time, the first prize went to Peugeot, who used a four-seater car.
In 1895 both Panhard and Peugeot had used the same engine, the new Phénix made by the former, which was an actual vertical engine instead of the Daimler narrow V-twin, but in 1898 Peugeot could resist temptation no longer, and the cars that took part in the race from Paris to Marseilles and back were fitted with the new Peugeot horizontal engine. In this, the cylinders were parallel, with their heads pointing backwards, and both cranks were on the same throw. The motives underlying this change may be read between the lines of Gerald Rose’s comment. “As the plane of rotation of the flywheel was the same as that of the driving wheels of the car,” he says, “all bevel gears were eliminated, and the drive was through an intermediate shaft, carrying the train of gearwheels for the various speeds, on to the differential shaft, which carried the sprockets for the side chains . . . “
For Peugeot, the decision was to prove a disastrous one, even if the fact was not immediately apparent. True, for the first time, the firm was decisively beaten by Panhard et Levassor, but this appeared at the time to be due to the insignificant, and easily rectifiable, fact that the burners for the tube ignition on the new engine were excessively exposed, and as the race was run in the September equinoctial gales, they incontinently and continuously blew out. Moreover, the horizontal engine school was growing rapidly in numbers and prestige. Apart from the usual swarm of Benz derivatives, masquerading as Parisiennes, Fissons and Trionleyres, there were the new Delahayes, described as “perhaps the most original and improved cars in the race,” and the new Amédée Bollée quadricycle. The Delahaye, actually, was basically on Benz lines, with its two-cylinder horizontal engine at the back, driving forward by belts to the countershaft, and back to the rear wheels by side chains. But the Amédée Bollée quadricycle was a monument of misplaced ingenuity. Bollée agreed with Levassor that the engine should be located at the front, but having placed his two-cylinder motor there, he arranged it horizontally, with the crankshaft running across the car. Next he seems to have been worried about keeping the back wheels on the ground, with the result that he placed his change-speed gear behind the back axle and drove to the countershaft with an immensely long belt. Everything was still revolving in the same plane as the road wheels, but instead of using chains for his final drive, he stultified his whole arrangement to date by patting a bevel gear on each end of his second motion-shaft, gearing with another on the end of a universally-jointed longitudinal shaft, running forward and having another bevel on its forward end, gearing with an internally-toothed ring on the road wheels. Amédée Bollée, in fact, obviously did not object to bevels—instead of putting the starting handle in at the side with a dog clutch on to the crankshaft, he put it in at the front, with a bevel gear on the end of it. Indeed, his predilection for the horizontal engine may have been merely due to the fact that, in the words of Louis Lockert, a contemporary authority, it “does away with the intolerable vibration experienced in cars in which the motors are placed vertically.” In any case, by the end of 1896, Emile Levassor, as a perpetrator of such “intolerable vibrations,” was more than ever alone in his infamy.
The events of 1897 did little to settle one way or the other this fundamental argument as to the best “system” on which to construct an automobile. The year is remarkable in that, while several races were organised, it brought forth no classic event worthy to rank with Paris-Bordeaux-Paris and Paris-Marseilles-Paris. Marseilles-Nice-la Turbie at the beginning of the year was won by a de Dion Steam brake, but among the petrol cars, Lemaître’s Peugeot succeeded in beating the Panhards. On the other hand, in Paris-Dieppe and Paris-Trouville, later in the year, Panhard managed to beat Peugeot, although in both races there was mighty little in it. The year, too, saw the appearance of Mors in the arena, but at this stage Mors, or his designer Brasier, was a “trimmer” in the vertical versus horizontal engine controversy, the four-cylinders of the Mors engine being arranged in a V, and inclined at 45 degrees both to the vertical and the horizontal. In so far as he placed the engine at the back, with its crankshaft across the car, however, Brasier sided with the opposition against Levassor—or rather against his mechanical heirs, the great man himself having died on April 14th.
Up to now, practically all designers had contented themselves with two-cylinders, although in Paris-Marseilles-Paris the winning Panhard had had four. In 1898, however, after Levassor’s death, his ghost, so to speak, played its trump card. With the longitudinal crankshaft arrangement, it was far easier to increase the number of cylinders, and therefore the length of the crankshaft, than it was if the engine was set across the frame. In the Paris-Amsterdam-Paris race, all the new Panhards had the 8-h.p. four-cylinder engine, while their rivals carried on with their twins. As a result, one must imagine, the Peugeots, for the first time, were not in the picture. Indeed the sternest opposition came from the Bollées; and Panhard finally occupied the first two places, after one of the team had been in the lead at the end of every one of the six stages.
In 1899 Peugeot replied with a huge two-cylinder engine, having a bore and stroke of 140 by 190 mm., and a capacity of nearly six litres, armed with which Lemaître, in the Nice-Castellane-Nice race, succeeded in beating the 8-h.p. Panhards, with their comparatively diminutive 2½-litre four-cylinder engines. For some reason, however, this Peugeot did not take part in the later races of the year, and when it did reappear, in the Paris-Toulouse-Paris race of 1900, it did no good at all. By adopting the horizontal engine, the house of Peugeot, once the peer of Panhard, had ruled itself out from the great races, in which for some time thereafter it took little further part. After a decent interval, its designers ate humble pie and designed a voiturette with a vertical single-cylinder engine in front; and having thus gone down to the bottom of the class, proceeded to work themselves up again to an eminence as lofty, perhaps, as any previously occupied by their old rival. All that, however, still lay far in the future.
In the meantime, the Tour de France of 1899 had been largely a repetition of Paris-Amsterdam-Paris of the previous year. Panhards dominated the race throughout, with the Bollées still their only possible rivals, though by now obviously less formidable, and finally occupied the first four places in the race. The heirs of Levassor were at the apogee of their supremacy. Levassor himself, however, was dead, while Brasier, the Mors designer, was very much alive; much too alive, in fact, to allow the obvious for long to escape his attention. In 1899 there appeared a new Mors racing car, which, in complete contrast to its predecessors, had a vertical, “longitudinal” engine in front and in all other essential particulars followed the Levassor system. By 1900, moreover, it was apparent that the design incorporated certain improvements on the contemporary Panhard. The latter’s engine dimensions had by now grown to 110 by 140 mm., but those of the Mors, at 119 by 165 mm., were larger still; and with its low-tension magneto, instead of combined platinum tube and coil ignition, the Mors could probably make better use of its capacity. In 1900, Levegli on the 24-h.p. Mors, won Paris-Toulouse-Paris, and for the first time Panhard was defeated in the great race of the year. For the first time, since the defection of Peugeot, la marque doyenne had been confronted with a rival who used a vertical engine.
Nineteen hundred and one saw the appearance of the Mercédès, in which certain observers in recent years have sought to discern a revolutionary machine, and the only true ancestor of the modern motor-car. Looked at in its proper historical perspective, however, the Mercédès, I think, clearly appears, like the Mors, as a Panhard with improvements—even if the improvements in this case were marked. The break which it revealed in the Cannstalt tradition, however, demonstrated how strongly the tide was now running in favour of Levassor’s principles. While Panhard, Mors and Mercédès dominated the big racing car class, Darracq and Renault, with cars of basically similar lay-out, were equally supreme in the classes for light cars and voiturettes. De Dietrich abandoned the Amédée Bollée design, and adopted another, on Levassor principles, by Turcat-Méry. Even Benz had to submit, and call in a Frenchman, Marius Barbarou, to build them a car on French lines. When in 1904 F.I.A.T. made its belated appearance on behalf of Italy, it had abandoned its old rear-engine lay-out in favour of a Mercédès simulacrum—and thereafter successive Italian designers simulated F.I.A.T. By this time every single serious Continental designer of racing cars was subscribing to the principles laid down by Levassor—and by Levassor alone—ten years before.
But the history of the horizontal engine was not closed. Indeed, that aspect of its history on which I intended to concentrate when I began writing this article has not yet been touched on. You are warned, therefore, that “Sideslips” is about to take on a serial form: await the next thrilling instalment in an early issue!