(first posted 11/17/2014. In the five years since this was first posted there have been a number of updates. The year of the Premier was originalyl given as 1903, but has been changed to 1905 on the basis of more accurate information than the Indianapolis Speedway’s sign for it, which has since been updated too. I have continued to research the issue of who was the first to use a hemi head with canted valves, and there’s growing evidence as to where and when that happened, but it’s not absolutely conclusive yet. But it’s still my goal to eventually determine the true origin of the hemi head)
This 1905 Premier racing car was the first automobile to utilize a hemi head engine, and with an overhead cam at that. Although its actual racing career was truncated due to a misunderstanding of the rules, and it only raced once, its influence was outsize, even in relation to its giant 923 cubic inch four cylinder engine. Within a few years, just about every successful top-tier racing car sported an OHC hemi head or a variation of it, as would eventually most all modern automobile engines. Thus it deserves to be recognized as one of the greatest automotive milestones.
But the Premier’s designer didn’t invent the hemi head, he saw it in use on a boat. And although at least one individual has tried to take credit for first conceiving that hemi marine engine, it’s not a credible claim. But someone on the shores of Lake Michigan in the final years of the 19th century did first conceive of the hemi, and a name has been put forward. Is he the true father of the hemi?
Tracking down the origin of any specific automotive technological development can be challenging, especially if it occurred in the mists of the late 19th century. When exactly did it first occur to someone that the key to greater engine performance lay in its breathing, and thus in its cylinder head configuration, valve size and porting? The hemi head, with its large canted valves and unobstructed cross-flow porting, was the obvious and optimum solution, and the definitive breakthrough to all subsequent higher performance gasoline engines. Inevitably, someone was going to figure that out, but who was the first?
Before we delve into the murky origins of the hemi, let’s examine the hard steel facts in front of us. The Premier is a very significant piece of automotive history, given its pioneering engine. But it’s not been widely credited, even at this museum where it sits. It’s time to give it its justly deserved recognition.
I first encountered it in this print of a splendid painting by Peter Helck, called “Death Defying Combat”, in an Automobile Quarterly many years ago. In addition to its dramatic subject and style, I was struck by the huge air-cooled engine with its exposed OHC hemi valve train in that utterly bare-bones racing car. Automobile Quarterly provided some info on the painting, but nothing about the Premier’s cylinder head, which rather engaged my hemispheres.
According to AQ, this event never occurred. It’s an imaginative depiction of a grudge race that could have taken place between two of the most notorious racers in 1905. The “Green Dragon” (on right) piloted by the famous Barney Oldfield is a 60 hp Peerless that held numerous records on the borrowed horse-racing dirt tracks of the time. The Premier, sporting a then-massive 100 hp, had been built specifically with the intention of challenging Europe’s finest racers in the 1905 Vanderbilt Cup Series, a road race on Long Island, NY., with a valuable prize. It was a precursor to Grand Prix racing, and Americans stepped up to challenge the Europeans with their fastest machines.
But somehow, the memo restricting Vanderbilt Cup contenders to a maximum weight of 1000 kg (2204 lbs) was never received by its owner Carl Fisher, because the Premier’s huge 923 CID engine pushed it some 300lbs beyond that limit, despite the exposed crankcase and no exhaust pipes to speak of. In a desperate move, Fisher had the chassis and engine drilled out as much as possible (470 holes), and replaced the heavy bevel-gear rear axle with a lighter chain drive, but all to no avail. It was still some 80-120 lbs overweight, depending on the source. Fisher and Wideley requested an exemption with the understanding that the Premier would lose the additional weight before the race, but they were turned down.
It would be interesting to speculate how the Premier would have made out in the grueling 284 mile long 1905 Vanderbilt Cup, which was won by a French Panhard averaging 52.2 mph. In this shot, Louis Chevrolet in a FIAT is challenging the winner George Heath in the Panhard.
Not surprisingly, Fisher was not happy about his treatment, and made his feelings quite clear in this paid ad.
It’s known that George Weidley designed the Premier racer, built by the Premier Automobile Co, of Indianapolis, but despite the brilliance of its design and execution, Weidley is not the hemi-father we’re looking for. Weidley’s brief from Fisher was to build the fastest racing car in the world, and what a monster of an engine it is: four cylinders with a 7″ bore and 6″ stroke, displacing 923.43 cubic inches (15 Liters), making 100 hp at 1000 rpm, a superlative output for that time. Torque? Perhaps no device strong enough existed to measure it.
One thing it doesn’t have: an enclosed crankshaft: yes, those are the connecting rods down there, acting on a fully exposed crankshaft. The driver and mechanic undoubtedly were well-oiled, if not the bearings. And in front, that’s the bevel gear for the shaft to drive the overhead camshaft. Lubrication? Most of these early engines had total-loss oiling systems, with a large oil tank that was pressurized by a hand pump, which can be seen between the seats. It was the riding mechanic’s job to deliver the appropriate amounts of oil. Crude, but the oil that was delivered was always fresh and clean, at least. The typical small external brass oil lines are nowhere to be seen on this engine; presumably they were removed at some point.
The Premier only raced once, in a five-mile handicap race at the Indiana State Fair, where it handily won with Mr. Fisher at the wheel, and timed at 59.21 mph (99 kmh). Or is that painting actually of that race? There’s a fair number of conflicting “facts” and memories on these distant events.
Here’s a shot of the legendary Barney Oldfield behind the wheel in Chicago in 1906.
We do know that Mr. Fisher went on to build his own little race track: Indianapolis. But although the monster he brought to life never raced again, it was brought out for an early vintage event at the brickyard. Note the lubrication system (now missing) with its central reservoir and numerous pipes leading to the various point on the engine. This type of “total loss” system was typical in the early days, and the driver had to remember to repeatedly push a plunger or pull a handle to send oil to the engine.
I unexpectedly encountered the Premier at the Indy 500 Museum, and running into this beast was the highlight of the visit. And it led to my deep dive into the origins of this remarkably advanced cylinder head design. But before we do that, lets explore some alternate very early hemis, which may or may not have been influenced by the Premier, complicating the unraveling of this story.
The brilliant Austrian engineer Hans Ledwinka, soon to become famous for his many innovations at Tatra, drew up a rather similar SOHC hemi head engine sometime during the time after he returned to Nesselsdorfer in 1905, which resulted in the NW Type S, in both four (S4 – 3.3 L) and six cylinder (S6 – 5.0 L) versions. But since these did not go into production until 1910, it’s an open question as to just when Ledwinka gave it a hemi head. But apparently prototypes were already built as early as 1907 or 1908.
These cars were fast for the day (100 kmh/60 mph for the S6) and highly advanced, thanks to their exceptionally powerful engines for their displacement. They generated some 10hp/L, a superlative result for the time and almost twice what the Premier made (6.6 hp/L). With the Model S, Ledwinka not only saved Nesselsdorf, which later became Tatra, but also made his first lasting mark on the automotive world.
Including Ferdinand Porsche. When his Austro Daimlers proved themselves to be too slow in the 1909 Prince Heinrich Trial, he created a radically advanced car for the next year, one that utterly dominated the Trial in a 1-2-3 finish. The 1910 Austro Daimler 22/86 “Prince Heinrich” has been christened “the first true sports car”, because it was not a racing car, but a perfectly streetable touring car capable of carrying four (as required in the trial) at up to 140 km/h (87 mph). Porsche himself drove the winning car (above).
There were two key technological advances to its overwhelming superiority. One was the first thorough application of aerodynamics. Although it may not look that way, every aspect of its design was streamlined, given the basic configuration of the day. Even the axles had a tear drop form, as did the tiny headlights. The body was fluted to minimize frontal area, and had a tapered end cap and full underbody cladding. Porsche was one of the first to realize that aerodynamic drag was the greatest single obstacle to speed. The other was power.
And the 1910 Austro Daimler had an abundance of that, thanks to Porsche’s splendid new OHC hemi engine. He ditched the previous T-head flathead, and designed a new 5.7 L four cylinder with large valves and big, smooth ports. This engine has all the requisite qualities of the modern high output engine.
Porsche and Ledwinka were friends, and regularly discussed their latest work, or just simply looked over each other’s shoulders from time to time. Although Ledwinka’s S-Type did not go into production until 1910, protoypes had been built a couple of years earlier. And since Porsche’s 1909 Prince Heinrich racer used a T-head, it’s pretty safe to assume that he was likely inspired by Ledwinka’s hemi head. Porsche’s new OHC hemi configuration appears to be a direct evolution of Ledwinka’s.
Lest Porsche be accused of being an empty-suit plagiarizer, inevitably Ledwinka’s advanced engine was going to be copied; Porsche presumably was just the first. And it wasn’t a blatant copy; he developed and refined it in his usual thorough and compulsive way, and tested it himself relentlessly until it was just right. It was significantly different in many ways, but the inspiration was clearly Ledwinka’s NW Model S.
Porsche’s changes and attention to detail led to a powerful but reliable engine that produced 95 hp at 2100 rpm, and 238 ft.lbs of torque at 2000 rpm from its 5.7 liters. Despite being perfectly tractable and suitable as a regular road car, this was the highest specific output per liter of any engine at the time (16.66 hp/L). The hemi head had truly arrived, and the 1910 AD Prince Heinrich has gone down in history as one of the most significant cars ever.
Before we stop digressing and go back in time in search for the hemi’s origins, let’s just briefly acknowledge that Ledwinka’s and Porsche’s SOHC hemis were soon copied widely in Europe, and the next step of the hemi’s evolution arrived within just a few more years. That was the 1912 Peugeot L76 racing engine (above), the first with a DOHC hemi/pentroof head and four valves per cylinder (which essentially requires the head shape be more pent-roof than true hemispherical, but the same principles apply). The Peugeot dominated Grand Prix, won the 1913 Indy 500, and was universally copied. The mark it left at Indianapolis was especially profound: it was the basis of the fabulous Miller and Offenhauser engines that dominated Indy for so many decades thereafter. And of course, DOHC four-valve engines are ubiquitous now, even in the lowliest and cheapest cars.
The very big question is just what inspired Ledwinka’s SOHC hemi head? There’s little or no doubt that he didn’t just dream this up from scratch. It’s time to look closer at the origins of the hemi.
There were some previous attempts at combustion chambers approaching a hemi head in Europe. The European hemi origin myth gives credit to Boris Loutsky for developing a winning large marine racing engine built in Germany in 1902 by Daimler and designed by William Maybach. Clearly it’s not a hemi head in the classical configuration, and Ludvigsen refers to it as a “truncated cone”. Although it’s not a true hemi, and not optimized for most efficient gas flow, it clearly is a significant early effort at addressing the issue of improved breathing through larger valves and improved porting. And the basic configuration of the single center ohc and rocker arms operating the canted valves looks very similar to what Ledwinka used on his 1905 NW S, but relocated to the top-center of the head, which was now a true hemispherical head. The similarities in their valve trains is unmistakable.
The question as to whether Ledwinka saw the potential of the Loutsky design and improved it, or whether he somehow became aware of the 1905 Premier’s hemi head cannot be answered with certainty. Ludvigsen says it was “unlikely”. The Premier never raced with the Europeans at the Vanderbilt Cup; if it had, quite likely its head would have drawn attention and imitation. Also, the configuration of the location of the cam and rockers on Ledwinka’s head is much more similar to the Loutsky/Maybach/Daimler configuration. So Ludvigsen’s assumption is probably right, although it’s not certain.
And it’s certainly not that great of a coincidence that Ledwinka would synthesize essentially the same ohc hemi head thousands of miles away, given that it was intrinsically the optimum solution. Someone would have done so anyway, probably sooner rather than later. As such, it is another feather in Ledwinka’s cap, to which he would add so many others.
Time to head back to the US, and to the Truscott Launch and Engine Co. in St. Joseph, MI.on the banks of Lake Michigan. It’s been well documented that a series of ohc hemi head marine engines were in its catalog, starting in 1901, and a few of these engines are still in existence and working. But who came up with it?
This site shows a restored example of a two cylinder engine and credits “HEMI inventor Allie Ray Welch, who also founded the Welch Car Company, in 1901, for designing it. The text says “though it performed better than his other engines, the open (exposed) valves threw oil on the passengers, and for that reason, it was discontinued.” Well, that’s a start, and the part about the oil flying from the exposed valves is undoubtedly true, but it’s hardly the full story.
The confusion arises because Allie Ray Welch did file this patent for the OHC hemi head in 1905. And there is reference to the engine in the 1902 Welch automobile having such a head design. But that is easily debunked.
Welch claimed his 1902 automobile engine (above) had a “hemispheric combustion chamber” but it’s hardly a genuine hemi, and the valve arrangement is rather odd. It had two intakes and one exhaust; the two intakes on the side were atmospheric (no cam actuation), and the single exhaust is on top, mechanically activated by something described as a “disc”. The point is that although the Welch engine shows some understanding of the benefits of larger valves and better breathing, this is not a true or classic hemi head.
Welch did go on to build a true hemi head engine in his cars, starting in 1905, and they were unusually powerful for the times. And the valve train was essentially the same as used by Ledwinka, with a high cam and rocker arms. But why did Welch drop the earlier head design and adopt the one from the Truscott boat engine?
Almost certainly not, since Truscott had been building hemis since 1901 (this is a 1902 heavy duty unit from their catalog, note the fully-counter-weighted crankshaft). There were associations between Welch and Truscott, which cast the engines for Welch in its foundry. Undoubtedly Welch encountered the Truscott hemi there, and saw that it was a better solution than his own three-valve “bathtub” head. And he went ahead and filed a patent on it.
Not that there’s any mention of that very important historical milestone in its placard at the Indy 500 Museum. And even the year is wrong. Pretty pathetic, given that Fisher was the founder of the Speedway. Note: the sign has been updated to “1905”.
The final question is who did design the first Truscott hemi?
I came across a rather fascinating post and comment thread at theoldmotor.com from 2011 about the 1902 Truscott engine and the source of its inspiration. Several expert commenters weighed in on the subject, but none had a defining theory. But the very last comment, added two months later, caught my attention. A David Gierke left this tantalizing comment:
In terms of the origin of the hemi engine, inventor, Augustus M. Herring experimented with the hemi while he worked at Truscott’s Boat Manufacturing Co., in St. Joseph, Michigan in 1898. I have much more if anyone is interested.
I contacted Gierke, and we have exchanged a number of emails. The person he was referring to was Augustus Herring, one of the pioneers of flight. Herring (like the Wrights) knew that the keys to heavier than air flight lay in controlling the airship as well as an engine powerful enough to provide sufficient thrust.
In 2018, Gierke published a two-part biographical novel of Augustus Herring, which I reviewed here. It’s extremely detailed, including Herring’s work with engines. And although Gierke claims that Herring essentially “invented” the hemi head engine, there’s no definitive proof, at least in terms of the hemi head engine in the modern sense, with two angled valves.
Gierke does offer this drawing (by the author) of Herring’s 1902 model aeroplane engine. There’s at least four problems with this being the “smoking gun”: 1.) it’s not a typical four-stroke engine, as it has exhaust ports in the lower portion of the cylinder. 2.) it uses an atmospheric intake valve. 3.) the drawing does not show how the porting of this engine actually works. There’s no connection between the valves and the exhaust stack and intake/carburetor. 4.) It’s from 1902, which is a year after the Truscott marine engine was in production, which had all the hallmarks of the definitive hemi head engine.
But the connection between Herring and Truscott was very deep, inasmuch as Herring’s shop was a rented area within Truscott’s factory, which burned down in 1898. Given that Herring was hard at work in developing a more powerful IC engine for his airplanes and his “Mobike”, which was a pioneer motorcycle from this period (1898-199), and there was interaction between Herring and Truscott, it simply can’t be a coincidence that an old-school boat builder would suddenly “invent” the modern hemi-head engine out of the blue.
But so far Gierke has not been able to demonstrate with any drawings, documents or otherwise that Herring specifically designed that first Truscott hemi-head marine engine, or had some defined role in it. None of his other engines show the use of those distinctive characteristics up to that time, or even some years later, at least to the extent that they can be documented. Gierke claims that Truscott “ripped off Herring’s hemi head design”, but can’t show the evidence that Herring had one at or before that time.
Even if we can’t verify the true father of the hemi yet, George Widely recognized its potential when came across it, and his 1905 Premier’s engine has secured its very special place in history. Now if I could just hear it run, full chat, at its maximum engine speed of 1000 rpm. With a load, and from the driver’s seat. ‘Death Defying Combat’, indeed. Or hopefully so.