Veteran to classic
Was it built?
It is well known to Alvis enthusiasts that the design of the first engine for these cars, the 10/30, was the work of a Mr GPH. de Freville who also invented the name Alvis. It was de Freville an Englishman, who sold his engine design to TG John, founder of Alvis Ltd, around the year 1918. He had been Manager of the Hanover Street, London, showrooms of the DFP Concessionaires when Bentley and Bentley (WO. and his brother HM) took them over, and the engine which became the Alvis power unit was thought to have been based on all that was best in the DFP light-cars, of which WO Bentley also approved, having raced them, and taken records with them, before the war (Birkin also started his post-war racing career with a DFP at Brooklands). After the 10/30 Alvis had been put into production in Coventry de Freville drove one in the 1920 MCC London-Edinburgh Trial and entered, that year for Shelsley Walsh but was a non-starter. He had no more to do with Alvis but the interesting thing is that before the autumn of 1921 he had put on paper a notably advanced lightcar engine. This had dimensions of 63.5 x 118.5 mm (just under 11/2 litres); which compares with those of the 10/30 Alvis engine of 65x 110 mm, showing that de Freville favoured the long piston stroke. His new engine design incorporated four valves per cylinder in a detachable cylinder head these valves inclined to form hemispherical combustion chambers with the sparking-plugs centrally disposed. Operation was by a single overhead-camshaft driven from the front by a chain of gears. From these gears a skew-gear drive turned a cross-shaft from which two plunger oil-pumps were actuated, one a scavenge-pump, the other a pressure-pump, for dry-sump lubrication. '
Two advanced aspects of this de Freville engine were wet cylinder-liners ·that were screwed into the detachable head and slipped into the cylinder block, sealed at the base with copper washers, and the fact that the cylinder block was specified to be made in cast aluminium. Another unusual feature was the use of a built-up crankshaft so that a centre ball-bearing could be used, in conjunction with two more ball-bearings at the ends of the crankshaft. The big-ends were plain white-metal bearings and the cooling system incorporated a water-pump driven from the timing-gear train.
The use of an aluminium cylinder block was perhaps no surprise, because de Freville had pioneered light-alloy pistons for the Alvis engine, having seen these used by DFP before the war, as had WO Bentley, although de Freville' maintained that before that they had been used by Chenard-Walcker (and some historians say as early as 1906, by Aquila-Itala). So he was used to the casting techniques. A few speculations arise after studying this 1921 de Freville design for a high-performance light-car engine, apart from speculation as to whether one was ever built. One source, since dismissed as erroneous, had it that de Freville had had associations with Bugatti in France. While I would hate to disagree with those historians who say this is nonsense, if there was little in common between the Alvis &nd Bugatti engines, the link might be called closer when it came to the 16-valve, ohc, ballbearing de Freville engine of 1921 . De Freville had owned a foundry in Wandsworth during the 1914/18 war, and had supplied many of the aluminium castings for the Puma aero-engines made by Siddeley-Deasy in Coventry; where TG John was then the works manager. Perhaps it was from this foundry that de Freville intended parts for his new engine would be made, especially the cast-alloy cylinder block!
One wonders, also, how much the first 3-litre Bentley engine influenced de Freville's 1921 light-car power unit? - overhead-camshaft, albeit differently driven, multiple valves, dry-sump lubrication, long piston-stroke. After all, he had known WO very well before the war and would no doubt have followed his line of thinking thereafter.
All speculation, of course. It seems that de Freville went to America in 1924, as designer at the Lafayette Motor Company at Indianapolis and in later life concerned himself with space-travel matters. I do not think Lafayette would have wanted to 'develop the advanced 11/2-litre engine he had put on his drawing board. WB.
So what was it?
Doing some further research on JG Parry Thomas, my favourite racing motorist, I chanced upon an item which seemed to me rather interesting. Thomas took his straight-eight Leylands to Boulogne for the annual speed-trial and speed hill-climb, in 1923, 1924, 1925 and again in 1926. In 1923 Thomas made ftd in the aggregate for the two sprints and the hill climb but was disqualified for having changed the Leyland's tyres after the first three-kilometre flying-start run, after Rigal, who beat him by 0.05 sec in the 5-litre Panhard-Levassor, had put in a protest. Thomas's excuse was that he wanted to save the tyres for Brooklands record bids ...
In 1924 Thomas won, on aggregate, from Major Cae's 30/98 and JA Joyce in the little AC, the Leyland Thomas covering the ss three-kilometre course at a rousing 127.57 mph. In 1925 Thomas skidded in practice on the wet road and was extremely lucky to escape with only minor injuries, for the Leyland Thomas was badly damaged, as has been described and illustrated in Motor Sport in recent times.
The 1926 Boulogne speed week was abandoned after RB Howey's fatal accident in the hillclimb while driving his Ballot, but before this Parry Thomas been beaten only by Seagrave in the 4-litre V12 LSR Sunbeam, in the six-kilometre fs speed trial, run from Boulogne towards St Orner. Of this run, with the Sunbeam doing around 140 mph on the very narrow road, Segrave said he had never before been frightened in a motor car, but was definitely so on this occasion, Howey in the big Ballot was third. However it is not these fine performances by Thomas with which I am here concerned. The significant item I noticed was that in the 1923 Boulogne speed trial a La Perle had been placed fourth behind Thomas, Rigal, and Frank Clement in a 3-litre Bentley. With Thomas disqualified, his La Perle was third, giving away an aggregate 8.8 sec to the big Panhard and only a second to the Bentley. The significant point is that this was a 11/2-litre car. It was, moreover, driven by a gentleman called Du Fourmentin, who to my knowledge was not a well:-known racing driver, at all events outside France: Hang on, though! This driver in his 1,500 cc La Perle was a second quicker than Segrave in the 2-litre GP Sunbeam and a whole three seconds faster than Rigal on the 3-litre Panhard, who tied with the amateur Summers in his very fast 30/98 Vauxhall. Not only that, but behind this La Perle were 21 other drivers who included those of the calibre of Eyston (Aston-Martin), Benoist (Salmson), Coe in his 30/98, BS. Marshall in his Bugatti, and so on. And some of the cars that were easily vanquished by this little French racer were of anything from three-litres to five-litres ... so what was this amazing La Perle?
Very little seems to be known about this rare make, apart from it being the work of Louis Lefevre, at a factory in Boulogne itself. He first, just before the war, made cyclecars, like so many French experimenters, and after the German withdrawal he turned to light cars with Bignan engines. I think it was more likely that the car which did so well at the 1923 Boulogne event was a fore-runner of later La Perles, using the slightly-larger power-unit designed by M Causan, who was afterwards to provide La Perle with six-cylinder, single overhead-camshaft engines, still of 11/2-litres, and to build a supercharged racing edition. I have seen a photograph of a very sporting La Perle, looking absolutely ready to race, and perhaps better streamlined than, say, a stripped Type 37 Bugatti.
We might have known more about these things had the La Perle entered as a Pearl (a name I am sure Rolls and Royce would have been pleased to have used on another car had they thought of it!), not failed to appear when it was supposed to have run at Brooklands. That was in 1925. It had been entered by none other than Captain (later Sir) Malcolm Campbell, who knew a good car when he saw one, and was one of the sixcylinder jobs said to be painted saxe blue. But no engine dimensions were given apart from "under 1,500 cc", which was contrary to the BARC's entry list requirements, which makes one think that Campbell never actually took delivery of it, at all events before that Easter Monday Brooklands Meeting. Unless the time-keepers muddled things at Boulogne in 1923, it could have been a very potent little racing car. Had he started in his first race Campbell would have been sandwiched between Vernon Balls' Amilcar and Dr Benjafield's Salmson, both 1,100 cc cars, having to give the former three seconds, and the latter giving Campbell four seconds. So, although the handicappers did not usually give newcomers much chance, they do not seem to have been unduly hard on this 11/2-litre Pearl, in spite of that Boulogne reputation, which was perhaps unknown to them. Perhaps more erudite historians than I, with more time for research, can throw some light on just how potent the Perle was? - WB