Racing car evolution part III: 1922-1925

We have great pleasure in presenting herewith a continuation of the "Evolution of the Racing Car" series, this time contributed by Laurence Pomeroy, Technical Editor of "The Motor" and author of the "Milestones of Speed" articles now appearing in that journal. The first instalment, covering the period 1919 to 1921, appeared in our October issue. The text and illustrations are published by kind permission of Temple Press, Ltd., and remain their copyright. – Ed.

The period I am now about to review represents the zenith of motor racing from a viewpoint of the number of companies engaged. It is also a period of intense technical interest, particularly in respect of engine development, so much so that at the end of it Grand Prix engines were developing as much power per litre as they were 10 years subsequently.

So far as international racing was concerned a 2-litre capacity was enforced throughout, apart from three exceptions, which I think I shall deal with straight away. Early in the year the R.A.C. ran a T.T. race in the Isle of Man open to cars of 3-litre capacity; this attracted the previous year's Grand Prix Sunbeams, a team of Bentleys with tuned-up standard production engines, and three specially built Vauxhall cars with engines designed by Ricardo. In view of the obviously high costs to the last company of building special cars it may seem odd that they should have been chosen to enter a race under the virtually obsolete 3-litre formula and that they did not wait until the French G.P. and run 2-litre models which could have been kept going for some years.

The answer, believe it or not, was pure ignorance. I can vividly remember the astonishment of the Chief Engineer when I, then a very small schoolboy, told him the true facts, for he, up to that time, had been quite confident that the R.A.C. were complying with established international practice. As it happens, the Vauxhalls did not even receive a reward for their enterprise and could do no better than third, first place being secured by Chassagne on a Sunbeam, and the runnerup being a Bentley.

Indianapolis was the second 3-litre event and the straight-eight Ballots were sent over, one of them running third, the race being won by Jimmy Murphy on a Duesenberg chassis with a Miller engine. Another non-formula event was the Targa Florio, which was won by an ex-1914 G.P. Mercédès, which had been fitted with front-wheel brakes. Two Ballots ran second and third, but these were 2-litre sports cars having four cylinders, with two overhead camshafts; design by Henri.

The main race of the year was undoubtedly the French Grand Prix held at Strasbourg. Entries were Fiat, Bugatti, Ballot, Rolland Pialin and Sunbeam, also two Aston-Martins of 2 1/2-litre capacity. This race forms, as it were, the watershed between the pre- and the post-Henri school of design. The Aston-Martins had engines which were virtually half the previous Henri-designed 3-litre Ballots, the 2-litre Ballot was designed by Henri himself, with all the characteristics of his engines, and were of the type run in the 1921 French Grand Prix, gaining third place.

The Sunbeam was a new design by Henri, who had been retained by Louis Coatalen for this purpose. In this model the designer continued on the lines which had hitherto brought him and his copiers so much success, but rather surprisingly he abandoned the straight-eight engine in favour of a four-cylinder unit. This had an iron misting as cylinder block, three valves per cylinder inclined at a small angle. The crankshaft ran on ball bearings throughout with plain big-ends; the two camshafts were driven by a train of timing gears from the front end of the engine; the gearbox was built in one unit and the brakes had mechanical servo action. In view of the success of the straight-eight principle in the 1919-1920 period, it is strange that only two entries in 1922 had this number of cylinders, viz., the Rolland Pialin and the Btigatti. The former need not long detain us, as it was never a very successful design, although it embodied many ingenious mechanical features. The latter was a forerunner of a long range of successful cars during the decade. The crankshaft ran on three ball bearings, the cylinders were castings of two blocks of four, and single overhead camshafts operated two inlet, and one exhaust valve placed vertically in the cylinder head. Two carburetters were used, and it was claimed that the engine would give 90 h.p., although this is probably an exaggeration.

Both the Bugattis and the Ballots were fitted with egg-shaped bodies, and amongst the ingenious features of the latter was a scheme for taking the exhaust out through the centre of the body. It was supposed to favour streamlining, although it made the interior of the car extremely hot. The bore and stroke was 60 x 88 mm.

The Fiats had engines which were scaled down replicas of the 3-litre models I have previously described, the bore remained the same, 65 mm., and the stroke was reduced from 112 to 100 mm., giving a capacity of just under 2 litres. The general specification of the engine, viz., timing gears at the back, cylinder built from steel forgings with welded-on jackets, wide angle valves, and all-rollerbearing engine, was identical with the previous types, and the choice of six cylinders seems to have been dictated primarily by convenience and not by belief in the superiority of this number as compared to eight cylinders in a line. These Fiats developed 60 h.p. and ran up to 5,000 r.p.m. The piston speed at these revolutions was only 3,300 f.p.m., a direct contrast to the Sunbeam, for Henri's choice of a stroke-bore ratio gave him dimensions of 68 x 136 mm. so that 3,300 f.p.m. piston speed equalled only 3,800 r.p.m.

In a sense, therefore, the Sunbeams were beaten before they started, and, in fact, during the race at Strasbourg they were never in the same street as the Fiats. Their best lap speed was approximately 78 m.p.h., whilst the Bugattis were able to average 80 m.p.h., and the best Fiat speed was 86 m.p.h. Actually, Nazzara on a Fiat won easily at an average speed of 79.4 m.p.h., although rear axle troubles affected the other two at the last moment and prevented them having a grand slam. In consequence, two of the Bugattis finished second and third.

The superiority of the Fiats was so great that in the sole 2-litre event held during the year only the Bugatti came out to challenge it, and the Fiat had an exceedingly easy win on the Monza Circuit at the excellent average speed of 86 m.p.h. Thus, although Henri had really been represented by three makes in 1922, viz., Aston-Alartin, Ballot and Sunbeam, it was apparent that his basic designs were now superseded.

In 1923 the Fiat Bertarione school was in its first flush of success. The Targa Florio was poorly supported at the beginning of the year and was won by Alfa-Romeo with one of their large six-cylinder models, the first Alfa-Romeo victory in a major event.

Indianapolis was interesting because it was the last occasion on which there was strong European representation. Ettore Bugatti sent over a team of cars with Strasbourg type chassis and single-seater bodies, and Mercédès entered some four-cylinder models with bore and stroke of 75 x 129 mm., supercharged. These were successors of a rather unsuccessful blown 1 1/2-Iitre which had been entered by the Stuttgart firm for the 1922 Targa Florio.

Although the drivers were unfamiliar with Indianapolis they did tolerably well and one came home third despite exhaust valve trouble. They were by no means the first racing cars to be fitted with superchargers. Experiments in this direction had been made as long ago as 1907 and 1908 on an American car called the "Chadwick," whilst the Mercédès Company themselves had been experimenting with blowers since 1917, and made their first supercharged car in 1919. However, despite the winning of the race by an unsupercharged Miller at 90.95 m.p.h. (which was 4 m.p.h. slower than the previous year's record with the 3-litre model), the Mercédès entry focussed public interest on the blown type of engine.

Even more significant was the arrival of some entirely new Fiat models (with superchargers) on the starting line of the French Grand Prix held at Tours. It is doubtful whether there has ever been a race with more technical interest and diversity than this one. The Rolland Pialins turned up with the previous year's cars, but the other teams, Voisin, Bugatti, Fiat, Sunbeam and a single Delage were of extreme interest. Bugatti kept the same engine as previously, but had ultrashort wheelbase chassis with tank type streamlined coachwork, whilst the Voisin also had carefully calculated bodies of good aerodynamic form. Delage returned to racing with a car designed by M. Plancton and built in an extremely short space of time, orthodox so far as the chassis was concerned, but with a twelve-cylinder engine potentially capable of a very high r.p.m. and road speed, the stroke being only 80 mm., the bore 51.3 mm.

In the event none of these cars were serious contenders, and the struggle lay all the time between Sunbeam and Fiat. For the former Coatalen followed once again his policy of buying up the best brains from his rivals. During the previous winter he secured the services of Bertarione of the Fiats, and it is, therefore, not surprising the 1923 Sunbeam engines showed a very strong resemblance to the previous Turin products. In point of fact the 1923 Sunbeam engines were almost identical to that of the 1922 Fiat, except that 2 mm. were added to the bore and 6 mm. taken from the stroke, giving dimensions 67 x 94 mm., whilst the exhaust valve was made larger than the inlet valve. A unit construction gearbox giving only three forward speeds was fitted, and the external appearance of the car was again very "Fiat like."

The latter company had a team of cars of entirely new design by Messrs. Fornaca and Cavelli. They followed general principles previously established at Turin and, like the 1921 cars, had eight cylinders in line. The bore and stroke were 60 x 87.5 mm., and by moderate blowing the power was raised to 120 b.h.p., as compared to the 92 b.h.p. of the previous year's six-cylinder model and the 108 b.h.p. claimed for the unblown Sunbeam.

The contest between these two makes was close, for a Sunbeam was second at five laps, first at 10 laps, first at 20 laps, second at 30 laps, and Segrave finally won on the thirty-first lap. Fiats, on the other hand, were first at five laps, second at 10 laps, first at 20 laps and first at 30 laps. No Fiat finished, all of them going out with various mechanical troubles, including break-up of the rather primitive Vane type of blower, which will he described in detail in another part of this article. Sunbeam was second and fourth, Bugatti running third, 25 minutes behind the winner.

The Bugatti drivers were handicapped by the odd proportions of wheelbase and track chosen by their designer, viz., 6ft. 6 in. and 3ft. 3in. respectively, and this made them very hard to hold on the straight, although quick through the corners. At the time Bugatti attributed this characteristic (in my opinion falsely) to the streamlining, but the latter certainly proved its worth in respect of sheer performance. This was shown by some hand timing, which gave the winning Sunbeam a maximum speed of just over 130 m.p.h. and the Bugatti just under 114 m.p.h. although the Bugatti engine developed not more than 80% of the power realised on the Sunbeam.

This race was held in July, and by the time the Italian G.P. was held in September the Fiats had been fitted with Roots type blowers, with which they showed their quality in no mean fashion. Bordino led up to half distance, a staggering effort in view of the fact that he had broken his arm in practice and drove single-handed with a mechanic changing gear. He was forced to retire through physical exhaustion and Salamono won at 91.6 m.p.h., Nazzaro second and an unblown Miller, driven by Jimmy Murphy (the 1921 French G.P. winner), third, five minutes behind the winner. Fourth and fifth were two rear-engined Benz cars, designed under Rumpler patents. These cars had no influence at the time, but are worth mentioning because they led directly to the production in 1934 of rear-engined cars for Auto-Union, a story I must relate at some other time.

Fiats put up a lap record for the Monza track at 99.8 m.p.h., so the eight-cylinder blown cars thus proved themselves 6% faster over the whole distance and 7 1/2% faster over the lap. This was the first race won by a supercharged car, and since that time only one International G.P. has been won by an unsupercharged type, except on occasions where blowing has been proscribed by the regulations.

In 1924 the Targa Florio was won by a blown car, this being a four-cylinder Mercédès similar to the previous year's Indianapolis cars, but very much improved in detail by Dr. Porsche, who had become chief engineer. This was followed by Indianapolis, where Duesenberg appeared with an engine fitted with a centrifugal blower running at about six times engine speed and gave about 10 lb. boost. This car averaged 98.24 m.p.h. to win the event, and was thus 4 m.p.h. faster than the previous year's car, which had had one carburetter per cylinder.

For the circuit of Cremona, held early in the year, Alfa-Romeo entered some cars which had been designed for the 1923 Italian event, but which were withdrawn after an accident which killed one of their team, Sivocci. Winning easily at 83.9 m.p.h., Ascari was timed over 10 kilometres to do 123 m.p.h. This was clear proof that they would be formidable contenders for the French Grand Prix, which was run at Lyons on August 3rd. For this race the excellent entry of 22 cars was received, teams from Alfa-Romeo, Fiat, Delage, Sunbeam, Bugatti and Schmidt, and a single entry from the Miller. On these cars, Fiat and Delage were substantially the same as the previous year's models. The eight-cylinder Alfas were very "Fiat like" in general design, except that they had cylinder blocks in four pairs, but they were of an all-roller-bearing type, bore and stroke being 61 x 85 mm. Bugatti had much the same engine as previously, but a new chassis, the first of the famous Type 35 models, which are well known to all racing enthusiasts.

The Schmidt had cut valves and six cylinders and, with Sunbeam, were the only make having fewer than eight cylinders. In other words, after a lapse from favour in 1922 the straight-eight engine had returned to popularity in a big way. Although they were out of fashion in this respect, the Sunbeam cars were acknowledged the fastest on the course. They had the previous year's engines with Roots blowers added and developed 140 b.h.p. A rather longer and improved chassis included torque tube drive and four-speed gearboxes. They were clearly so superior to any other make that Alfas, who considered themselves the likely runners-up, approached the Sunbeam team manager with an offer that if he would permit Alfas to occupy second and third position they would gladly let Sunbeams take first place without pressing any of the cars too hard.

Needless to say, this offer was turned down and Sunbeams went to the line confident of victory, plus second and third. The night before the race they were visited by the Bosch Company, who exclaimed upon the burnt contact break-up covers of the magnetos, which were placed at the rear of the engine facing outwards towards the exhaust pipe. These particular instruments had remained untouched since the previous year, and no doubt the addition of the blower caused the piping to get much hotter and caused the burning to start. Bosch made a replacement with some magnetos of the latest type that had just arrived from Stuttgart. During the race the Sunbeams suffered from constant misfiring and, although superiority in speed was confirmed by unofficial timing of maxima during the race, which gave Sunbeam 130 m.p.h., Alfa-Romeo 124 m.p.h. and Delage, running unsupercharged, 114 m.p.h., nobody could account for the irregular running of the cars which constantly held them back, despite a record lap by Segrave. In the middle of the night following the race, their development engineer, Captain Jack Irving, leapt from his couch, rushed to the sheds, put back the old magnetos and the trouble vanished. By such a small incident were the Wolverhampton Company deprived of the rare honour of winning the French Grand Prix for two successive years! But Sunbeam ill-luck gave Alfa-Romeo the rare distinction of winning the French Grand Prix with their first entry for the event at 75.3 m.p.h. Micola Romeo and Ing. Jano must have felt very well satisfied. Seven years had to pass before they won this race again, although they had great success on other circuits.

The Bugattis suffered badly from tyre trouble and finished seventh and eighth, whilst the Fiats at the end of the tenth lap were running second, tenth, nineteenth and twentieth, a sad fall from grace after their predominance in the previous year. When Bordino retired with defective front brakes their four cars ceased to be effective challengers and, although Nazzaro survived until the twentieth lap before he also went out with brake trouble, he was then lying last.

In the Italian Grand Prix of 1924 no Fiats were entered, and four Alfa-Romeos were so much faster than any other car that when their slowest model finished, the next car, a Schmidt, was flagged off 10 laps behind. The winner, Ascari, put up the lap record to 104.24 m.p.h. and had a winning average of 98.76 m.p.h., i.e., he improved over the 1923 Fiat times by 8 1/2% and 5 1/2% respectively, and the former figure stood for seven years.

Just previously to this Sunbeams secured compensation for their French Grand Prix disappointment by winning the Spanish Grand Prix, Segrave corning in first at 62.54 m.p.h., a minute and a half ahead of a Bugatti, with a Delage running third. Bugatti put up the best lap at an average of 69.79 m.p.h.

In these two latter races a team of straight-eight Porsche-designed Mercédès were run, but although the engine developed about 170 b.h.p. the chassis design was so inferior that they had no success; in fact, I believe these cars never ran in a race without one of them having a serious accident.

In 1925 no new European cars appeared. Bugatti (still running unblown) won the Targa Florio. Then came the European Grand Prix, run by the Belgian Club on the Spa Circuit. In this race Delage appeared with blowers, new, two of; and connecting rods, old, 12 of. This combination proved unsatisfactory, the latter number being rapidly reduced. As a result, Alfa-Romeo were first and second.

For the French Grand Prix Bugatti had a modified chassis and engines with five main bearings, Sunbeams had the previous year's cars almost unchanged, Alfa ditto, whilst the Delage engine had been stiffened up inside and was developing about 190 h.p. at 7,000 r.p.m., as compared with 114 b.h.p. at 6,000 r.p.m. without blower the year previous.

Nevertheless, the Alfas got into the lead and held it until half distance, when the two remaining cars were withdrawn following a crash in which their leading driver, Ascari, was killed. At that time Delage were running second and fifth, with Sunbeam third. The finishing order was Delage first and second, Sunbeam third, Bugatti fourth, and it may be that Delage were playing a waiting game in the early stages of the race, for Divo was a little faster than Ascari for a lap and put in a record speed of 80.3 m.p.h.

Delage also won the Spanish Grand Prix at 76.4 m.p.h., whilst Alfa won the Italian Grand Prix at the rather surprisingly low average speed of 94.76 m.p.h., another Alfa being second and Bugatti third. Duesenberg ran fourth, and this make also put up the best lap of the race at 103.2 m.p.h., a little over 1 m.p.h. slower than the record put up for the previous year for the Monza Circuit. But at no time were Alfa challenged, and this accounts for the comparatively low speed recorded.

It will be seen that during 1925 Fiats had no part in European racing, but Bordino took a 1924 model to the U.S.A., and although he did not have any large number of wins he was timed to cover 25 miles on the Culver City board track at an average speed of 133.7 m.p.h. The American 2-litre cars, all now using straight-eights with centrifugal blowers and single-seater bodies on a light chassis, were even faster than this, and a Duesenberg averaged 139.7 m.p.h. for a timed distance. Indianapolis, won by the same car and the same driver (De Paolo) witnessed an average of 101.3 m.p.h., which stood as a track record until 1932. As time went on the Indianapolis and European road racing cars became increasingly specialised. The former became incapable of success on the road due to lack of progress in chassis parts, but so superior in attaining sheer speed that not until 1939 did a European car win at Indianapolis, after an interval of 20 years.

The twelve-cylinder Delage broke the flying kilometre record, timed in both directions at 134 m.p.h., whilst the straight-eight P2 model Alfa-Romeo went slightly faster than this and was later developed to average 138.15 m.p.h. over 10 kilometres. These speeds were not exceeded by any perceptible margin on road racing cars until the advent of the 1934 models, so we must give very high marks to the engine designers of A.D. 1925. Let us, therefore, examine in some detail the engineering aspects of the 2-litre formula.

A notable development during this period was a steady increase in weight, although the engine dimensions remained constant. In 1922 the winning Fiat had a wheelbase of 8 ft. 2 1/2 in. and weighed approximately 14 cwt. In 1923 the eight-cylinder Fiats weighed nearly 15 cwt., whilst the unsupercharged Sunbeams turned the scales at the exceedingly low figure of 13.3 cwt. By 1924, when Sunbeams developed new chassis to take the blown engine, having torque tube drive and a longer wheelbase (8 ft. 6 in.), the weight went up to 16 cwt., with Alfa-Romeo about half a cwt. lighter. Delage and Bugatti, running unsupercharged, were much lighter, being about 14 cwt. only.

Nevertheless, due to the greater power developed by the blown engines, the power-to-weight ratio figure very much favoured the latter type. In assessing this figure we should really take the car in running trim, i.e., with fuel and crew. If we do this we find that the 1922 Fiat weighed approximately 22 lb. per b.h.p., the 1924 Sunbeam 15 lb. per b.h.p. and the 1925 (driver only) Alfa-Romeo 12.5 lb. per b.h.p.; in other words, the power-to-weight ratio in three years increased by 77%.

The power per square foot of frontal area also rose from about 7 1/2 h.p. on the Fiat to 13 1/2 h.p. on the Sunbeam, to about 18 h.p. on the twelve-cylinder Delage of 1925. On this basis one would expect the maximum speeds of the cars to be of the order of Fiat, 100%; Sunbeam, 121%; Delage, 133%. As we know that the latter car had a timed maximum over the kilometre of 134 m.p.h., it follows that the calculated maxima would be:– Sunbeam, 124 m.p.h.; Fiat, 102..5 m.p.h.; and I think we may safely assume that under neutral conditions these figures are very close to the truth.

The whole of this increase in speed was due to improvements in dimension and design. I have set out the specifications of the leading motors in a table and have added as a matter of interest the 1922 Vauxhall with Ricardo engine, for although this was not a Grand Prix type it was designed in the same period, and being unsupercharged is strictly comparable with the Fiat of the same year and the Sunbeam of 1923. The figures realised enhance one's regret that the Vauxhall executive were not aware of the current racing regulations. It is clear that if they had done so and built a 2-litre model they would have had amongst the fastest cars in Europe.

A study of the table shows that power per litre, which is, after all, the determining issue under a capacity formula, was more than doubled during the course of the three years. The m.e.p. was raised by nearly 50% and power per square inch of piston area raised by 66%, and although piston speeds did not increase very greatly, r.p.m. were raised from approximately 5,000 to as high as 7,000 on the Delage by reducing the piston stroke. The strokebore ratios varied between 1.4 and 1.55:1, in place of the 1.72:1, which was so popular during the 1920-1921 period.

Compression ratios changed but little, being about 6:1 on the earlier engines and 7:1 on the 1925 types. Using atmospheric induction one would, from the foregoing considerations, expect maximum horse-power to increase from the 92 b.h.p. of the 1922 Fiat up to about 135 b.h.p. for the Delage, i.e., by 40%. It is, therefore, evident that the 100% gain in power per litre may be credited 40% to mechanical improvements and 60% to supercharging, and this in turn implies a supercharge pressure of about 7.5 to 9 lb. per square inch. The initial approach to blowing was much more tentative and moderate, and nothing approaching these figures was attempted.

As an example, whereas the 1925 straight-eight Mercédès had a directly driven and constantly engaged Roots blower, the first blown model from the Stuttgart works in 1922 had a blower engaged at will through a clutch and supplying air under pressure to the carburetter at 5 lb. per square inch. This practice was followed by the 1923 and 1924 Mercédès, and a scheme of similar intent, although different in function, was adopted on the 1923 Fiats and 1924 Alfa-Romeo cars.

On the former type the original installation consisted of a Vane type blower (patented by Wittig), in which there was a hinged flap surrounding part of the blades, movement of which increased or reduced the supercharger delivery. On the Alfa-Romeo a Roots blower was employed, but there was a by-pass valve allowing air to spill from the delivery to the intake side, thus once again giving control over the delivery pressure. On all these earlier blown cars the carburetter was placed under pressure, which meant, of course, a sealed float chamber and petrol tank with balanced pipes to both.

The blowers were not usefully employed at low engine r.p.m., for designers considered that to actually supercharge would cause detonation and/or breakage of gudgeon pins, piston crowns or bigends. The intention of the boost was restricted to restoring volumetric efficiency at high r.p.m. from the 75%, which would normally be a good figure, up to 100%. Actually, as the tables show, the first Fiat engines had no higher m.e.p. than the unsupercharged Sunbeam, and their slightly greater power was probably just as much due to their choice of eight cylinders instead of six as to the fitting of blowers.

This is probably because the early design of Wittig supercharger absorbed a good deal of power to drive it, and there is little doubt that when the Roots blower was installed for Monza the power output must have increased substantially.

The year 1924 really sees the beginnings of supercharging in a modern sense, for first Duesenberg at Indianapolis and then Sunbeam for Lyons adopted the now universal practice of aspirating through a carburetter and delivering a mixture to the cylinders at a pressure sufficiently high to give a fair measure of supercharge. With the continuously engaged blower applied thus, the increase in power and m.e.p. was held right down to low r.p.m.; as an example, on the Sunbeam the gain was approximately 25% at 2,500 r.p.m., 29% at 3,500 r.p.m. and 38% at 5,000 r.p.m.

One of the unexplained mysteries of the early years of motoring was the conservative ideas held by practically all designers regarding valve timing. Overlap between inlet and exhaust valve opening and closing has, on many modern engines been 70 degrees, or even more, usually divided equally on top dead centre. But on the 1922 Fiats the inlet valve opened on top dead centre and the exhaust valve closed only 10 degrees after t.d.c. Even on the 1924 supercharged Sunbeams inlet opening was only 9 degrees before t.d.c. and the exhaust 10 degrees after, a beggarly overlap of less than 20 degrees.

The period under review marks the end of petrol-engined racing cars, for as compression ratios were increased in 1922 and 1923 from 6:1 up to 7:1 it became necessary to add liberal percentages of benzole to the petrol. As supercharging became a fact as well as a name, it also became imperative to add alcohol.

It is not now possible to discover the exact fuel specifications used in this period, but it is likely that the earlier blown engines used something like a 40:60 petrol-benzole mixture, and that later types ran on about 40:40:20 petrol-benzole-ethyl alcohol.

In construction the V12 Delage stood alone, for the straight-eight type was dominant throughout. Design conformed very much to a pattern in which the crankshaft ran on ball or roller bearings and was of one piece, the connecting rods also having roller bearings and split big-ends and roller cages. Pressure lubrication through the crank was general, as were two valves per cylinder, inclined at a very wide angle around 100 degrees. In other words, the Bertarione concept replaced that of Henri.

Bugatti was consistently exceptional, with engines having three vertical valves per cylinder, two inlet and one exhaust, plain big-ends, and jet type lubrication to them. For the journals he employed a centre ball bearing on his three-bearing types, which had solid crankshafts, and five roller bearings on his latest models, which had built-up crankshafts held together by taper keys.

The cooling on the majority of the engines was exceedingly good, being facilitated by the wide angle of the valves; by contrast, Bugatti, with his vertical valves, found himself with what may be best termed iron-cooled head, although it must be admitted that performance did not seem to suffer greatly thereby.

Turning now to chassis design, one sees universal employment of four-speed gearboxes and typical ratios taken from the Sunbeam car give speeds for the indirect gears of about 93 m.p.h., 70 m.p.h. and 45 m.p.h.

Braking systems showed exceedingly little change, 14 1/2-in. brake drum, giving a brake area of about 1 sq. in. to every 5 lb. of car weight, being almost standard components. These were operated at high unit pressures between the brake lining and the drum, and this in turn involved a friction servo mechanism driven off the rear and off the gearboxes. Movement of the brake pedal applied a middle brake system running at one-thirtieth engine speed and the reaction was used to apply the normal brake shoes.

Frames, springs, shock absorbers and steering mechanism remained basically unchanged; in fact, the only chasms trend worthy of note is a gradual acceptance of the view that the rear springs should be relieved from torque either by torque tube drive or by separate radius arms, as employed by Bugatti. The latter designer, incidentally, was one of the few who omitted the brake servo motor.

Nevertheless, although chassis design to some extent stagnated, there is no denying that engines made greater strides in the two years 1924 and 1925 than they have done at any similar period subsequently. On twisty circuits, which put a premium on brake stability and cornering power, the 1925 2-litre cars were slower than the 1 1/2-litre types which followed them, but on really high-speed circuits they maintained their superiority for an astonishing length of time; it was, for instance, six years before the lap record at Monza put up by the P2 Alfa-Romeo in 1929 was beaten.

It was this disproportion between engine power and maximum speed on the one hand and inadequate chassis design on the other which led to the abandonment of the 2-litre formula in the interests of safety and the introduction of a capacity limit of 1 1/2-litres for the period 1926-27.

It is with a study of these years that I will finish these notes in a following issue.