Miscellaneous Internal-Combustion Engines

Opened: 31 Sept 2008

Updated 17 May 2018

Yet more on the Hewitt piston-valve engine


Back to Home PageBack to The MuseumBack to IC index

This gallery shows some of the truly bizarre engines that fit nowhere else; for example, the Polizzi flap engine, which looks like one of the great losers of all time.


This bizarre engine was shown at the Paris Flight Show in 1909.

Left: The Breton engine: 1909

The Breton engine propped up on a couple of elegant wood pillars. The object to the right is the stationary ignition magneto.

The Breton was a rotary engine; the engine rotated around a stationary shaft, and carried the propellor around with it. The peculiar design consisted of two V2 engines side by side. Three of these assemblies were mounted around a central shaft and drove it by spur gearing. Strangely the cylinders in the V2s were arranged to fire simultaneously, so the smooth flow of power theoretically possible with a 12-cylinder engine was thrown away. The cyinder heads for each set of four adjacent cylinders were a single casting.

The gearing was such that the three planetary crankshafts rotated at 1600 rpm, while the whole engine rotated en bloc at 400 rpm around the stationary central shaft, carrying around the propellor. It was supposed to give 60 HP at this speed.

From Flight magazine, 13 Nov 1909

It appears the two adjacent cylinders shared a combustion chamber, as only one spark plug per pair is shown in the photograph.

Flight magazine described it thus:

Dimensions: 82 mm by 85 mm, weight, 90 kilogs, price 10,000 Francs. The size is clearly far too small, and I suspect that some English journalist unfamiliar with the metric system confused mm with cm.

Left: The Breton engine: 1909

Air was drawn into the crank-case by fans on the three crankshafts for cooling purposes; one intake can be seen as a dark slot in bottom centre of the photograph. The cylinders were finned, but not deeply.

The valves were operated by cams on the central shaft, and fuel injected into the valve-chambers by pumps worked off the valve tappet rods. The curved injection pipes can be seen in the photograph above.

From Flight magazine, 13 Nov 1909

As far as I can determine this engine was never heard of again after the 1909 Paris Flight Show, but the idea of multiple V-engines geared to a central shaft resurfaced in 1939, with the Bakewell Wingfoot engine, shown below.


Left: The Polizzi flap engine: 1929.

Originator: Paul Polizzi, USA

Quite what benefit Mr Polizzi hoped to gain from this ludicrous arrangement is unclear. There would have been the usual sealing problems at the end and sides of the flap, plus extra difficulties sealing the pivot point at 12.

Info from Patent 62251


Having been rather rude about the the engine above, it came as rather a shock to find that there is a flapper engine under serious development today. Run, don't walk, to The Pivotal Engineering website. (Thanks to Jones The Engine for bringing this to my attention)

Left: Two, three and four chamber Pivotal engines.

These engines are assembled from modules.

Info from Pivotal Engineering website

Left: Two views of the water-cooled flap piston.

Info from Pivotal Engineering website

I only know what is posted on their website, but it does look as if they know what they're doing. They claim that their engine can have its combustion space thermally controlled, with no hot-spots, and that makes it especially suitable to run on hydrogen. Trying to run a normal IC engine on hydrogen results in catastrophic pre-ignition because hydrogen has such a low ignition temperature.


Left: The Bakewell engine front view: 1939.

The Bakewell engine looks at first like an ordinary radial, but it was nothing of the kind. It consisted of four 90-degree V-twin engines, each with its own crankshaft, which was then geared to a common central propellor shaft. Propellor reduction gearing was fitted, the standard ratio being 1.57 to 1. This form of construction was claimed by the makers to give "inherent balance".

The eight cylinders had a bore of 4in and a stroke of 4.5in, giving a displacement of 452 in3. (7407 cc) Output was claimed as 165 HP at 2800 rpm. The compression ratio was 5.4 to 1.

The idea of multiple V-engines had already appeared in 1909 in the Breton engine, shown above.

From Aerosphere 1939.

Left: The Bakewell engine rear view: 1939.

Fear me, earthling! The Bakewell engine was built by the Shaw-Palmer-Bakewell Co of Los Angeles. Why it was called "Wingfoot" is currently obscure.

Because of the 90-degree V angle, the cylinders nestling together in fact belonged to different V-twins.

At least one of these engines has been preserved.

From Aerosphere 1939


Left: The James engine: 1963.

This engine was put forward by Richard James Cylindrical Motors Ltd. It was a two-stroke engine with two swivelling opposed pistons operating on a single crankshaft at the bottom of the housing, and employing uniflow scavenging. It was based on work started as long ago as 1929. This is believed to be a picture of the prototype, which may well have been the only one made.

An friend of Alan Lea (who has kindly provided the report on which this article is partly based) said: "he'd seen one in Vancouver in the early 70's. He said he'd seen it running on an office desk with a (exhaust ) pipe stuck out of the window. No vibration or apparent unbalance."

The James engine apparently got some support from the Canadian Federal Government, but what happened after that is currently unknown.

Info from Setright

Left: The James engine: 1963.

The internals of the James engine.

The prototype engine had a swept volume of 29.2 in3 (478.5 cc) and a compression ratio of 11:1. The pistons were 2.375 by 2.750 in, at the rectangular section, and had a 2-inch stroke. (measured at the outer radius?) The engine body was of aluminium alloy with a cast-iron cylinder liner. The water jacket was cast integrally with the main block. The connecting rods were of drop-forged H-section steel, the big end having a single-row roller bearing; the small-end bearing were of the needle-roller type.

The pistons were sealed by L-shaped piston "rings" of cast-iron, pressed against the cylinder walls by spring expanders.

From report An Evaluation Of James Hemispherical Opposed Piston Engine by John J Szendrey Dipl. Ing., 17 March 1967. (Very kindly provided by Alan Lea)

James animation
Left: The James engine animated. Click on buttons to start/stop.

The fuel/air inlet is at extreme right. The transfer port is visible behind the right piston.

Another fine animation by Bill Todd. Javascript must be enabled for buttons to work.

Left: A James engine

This is claimed to be a 30 HP James engine. Dimensions are approx 12" wide, 14" high and 6" deep. It is mounted on a simple flanged stand.

This picture of a real James engine was sent to me by a correspondent; at the time he was hoping to sell it. It was given to him by John Szendrey, a mechanical engineer who worked on the engine at a Government research center in Vancouver sometime in the 1970ís. I am told James sold his patent to Peel industries, and John Szendrey got the engine working and was named by James as co-inventor. Note the mention of his name above.

Left: A James engine: other side

The cylindrical thing on the lower left is the starter motor.


Left: The Chater Lea face-cam engine: 1926

Chater Lea was a respected London-based firm which supplied frame fittings to bicycle and motorcycle makers, but also manufactured complete motorcycles. The single-cylinder 350cc face-cam engine appears to have been designed by Arthur Goodman and development engineer Dougal Marchant.

A vertical shaft driven off the crankshaft by bevel gears turned a pair of face cams, one above the other. The upper cam operate the inlet valve, and the lower cam the exhaust valve. A worm pump at the bottom of the vertical shaft fed oil up the hollow shaft centre to lubricate the cams; a separate pump supplied the crankshaft bearings and roller-bearing big-end. Note the spline connections to the shaft, presumably to allow for thermal expansion.

The possibility of desmodromic operation, (ie using the cams to positively close as well as open the valves, so valve springs could be dispensed with) was considered, but not adopted because it was felt that the mechanical clearances required for successful operation would be too critical for a non-technical owner to cope with. You did your own maintenance in those days.

The original source of this drawing is unknown, but from its style, was probably Motorcycling magazine.

Production ran from 1926 to the early 1930's, and approximately 550 face-cam models were made; this seems to demonstrate that it was a sound design. At least one example survives at The National Motorcycle Museum in Solihull.


This single-cylinder engine was designed by the motorcycle racing rider and manager Don O'Donovan, for Raleigh, who then made motorcycles as well as bicycles.

Left: The O'Donovan face-cam engine: 1930

The induction and exhaust valves were actuated by tappets bearing on a single horizontal face-cam. The vertical shaft to the face-cam was driven by bevel-gearing from the crankshaft.

The arrangement is very similar to the Chater-Lea engine described just above. According to Bird, O'Donovan "would have had ample opportunity to study it". Enough said, I think.

The O'Donovan engine was never actually used on a Raleigh but was sold to Dunelt who used it in one of their models.

There is unfortunately a terrible snag in this very neat and compact arrangement. Since the same cam is used for induction and exhaust valves, there is no opportunity for optimising perormance by having different event timing for the two valves. I Am Not A Motorcycle Engine Designer, but I would have thought this introduced crippling limitations. (The Chater Lea engine had a separate face-cam for each valve, and so did not suffer from this problem)

The original source of this drawing is unknown, but from its style probably Motorcycling magazine.

Info from "A Glimpse of the Vintage Years of Motorcycling at Brooklands" by Roger Bird. (self-published in 2008)

Left: The O'Donovan face-cam engine: 1930

Note the telescopic tube around the valve driveshaft, to allow for thermal expansion of the cylinder. The oil pump can be seen underneath the bevel box, exactly where it was placed on the Chater-Lea engine.

Left: Don O'Donovan contemplates valve-timings

He appears to have reached some unhappy conclusions.


The poppet valve has been the dominant valve technology for most of the history of the internal-combustion engine.
Other serious contenders have been the sleeve-valve and the
rotary valve.

However, the piston valve, so popular in steam engines, has always been a rarity, no doubt for good reasons. Here is an example that was patented in 1908 by James Mitchell Hewitt of Manchester.

Left: The Hewitt piston-valve engine: 1908

In Figure 1 the inlet valve h is on the left and the exhaust valve j on the right; both are driven by half-speed shafts geared to the crankshaft. The exhaust valve has water-cooling passages built in, as shown in Figures 2 and 3, and this is likely to be very necessary. The piston valves and main piston communicate with a large flattened combustion chamber. While I am not an IC engineer, I feel fairly confident in saying that this will give poor combustion and increased NOX output. I would be happy to hear from anyone with firmer information on this point.

The patent makes the intriguing point that the piston valves are not just valves; due to their piston action, with suitable timing they can help compress the charge and also will develop power on the expansion stroke. Whether this is either practical or useful I am not sure.

The big round thing on top is a mystery. It is not labelled in the drawings and is not mentioned in the patent. It may be something to do with the water cooling arrangements.

Apart from the patent, James Hewitt is unknown to Google, and I conclude the engine did not prosper, and was probably never built. The engine has nothing to do with the Hewitt Motor Company, founded in the USA in 1905 by Edward Ringwood Hewitt.

Image from British patent 25,411.
Application 25 Nov 1908, granted 25 Nov 1909.

Left: Hewitt obituary: 1918

James Hewitt may be unknown to Google, but not to Graces Guide. Here is his obituary.

I would like to know more about that railway accident. The Wikipedia list of British railway accidents has nothing for February 1918, and I fear that this might be the sort of accident where one person walks under a train. The Saver Clutch Company is also unknown to Google and this suggests it was not a great success.

The commercial success or otherwise of the Hewitt-Rhodes slipper brake is also currently unknown, but it can be said that US patent 714,798 was granted for it on the 2nd of December 1902.

From The Engineer for 8th March 1918

Paul Burke did a better job than me of searching for the accident that resulted in James Mitchell Hewitt's death. A good candidate is the derailment of the 1:45pm down express from Leeds to Carlisle, when it ran into a landslip. See here:


The actual accident occurred on 19th of January, but as well as 6 deaths at the scene, one further passenger later succumbed. The train was ultimately travelling to Glasgow, with which Hewitt was connected. The main argument against is that the coach worst affected was 3rd class, where Hewitt would have been most unlikely to be sitting. The report does not detail where the casualties were suffered.

Back to Home PageBack to The Museum EntranceBack to IC indexTop of this page