On the 11th hour of the 11th day of the 11th month, the “war to end all wars” ended. About six months later, a different war resumed, on American race tracks. Almost everyone else still battling for supremacy on big V-twins was racing stuff they’d pulled out of mothballs from almost three years ago. Not the Motor Company! Why and how this should be… is a pretty fair back story in itself.

You see, when World War I began for the U.S., all domestic manufacturers supplied motorcycles to our government for service in the conflict. Excelsior, Reading Standard, and the other small-potato builders sent a few, Harley sent a bunch, while Indian, in a fit of misguided greed disguised as patriotism, sold every motorcycle they made to the armed services. Two things wrong with that! First, it seems none at the Wigwam factored in the rapid, inflationary increase in costs of raw materials during a war. So the company made little or nothing on the deal and was broke when it ended. Second, since there were zero bikes made available for sale to civilians, dealers got pissed! Far more than a few jumped ship in order to make a living and supply what demand there was for those few miserable years. No award for guessing to whom most dealers defected. Harley-Davidson was the only major brand with bikes to sell to the public, for that whole time! A shrewd (not to mention profitable) decision on the part of the four founders. Now you can understand why Harley was so magnificently and maliciously set to become precisely what the founders always intended the company to be. By 1920, the manufacturer of the most and the best motorcycles in the entire world! That corporate strategy included domination of racing as well. The tactics, as usual, were left to Bill Ottaway. By the summer of 1919 the eight-valve racers were back to continue their winning ways. But they weren’t exactly the three-year-old stuff. No, sir! No 1916 race bikes and no moth balls in H-D’s arsenal… these things were new and different. Modern dyno testing on a couple of eight-valve racers show horsepower in the 60-plus range and a ready-to-race weight of about 250 pounds. No wonder they could hit 130 mph and manage 100 mph averages in races! Do the power-to-weight math… these century-old warhorses are probably quicker and faster than your new bagger! Ottaway did it… again! OK, full disclosure… different applies more than new. After all, the eight-valve was still the eight-valve, though eventually offered in four (or more) distinct varieties because Mr. Ottaway was what would be known now as a “development” engineer par excellence. Since the engines were used in several types of professional racing events, it stands to reason they would be tailored to suit each of them. So they were! So were the chassis… come to think of it. That’s where “development” naturally enters into it. History shows there have only been a handful of men who were superb at this task. Perhaps the best known example is Joe Craig, godfather of the fabled OHC Norton racing singles. To keep 500cc singles competitive for over 30 years is admirable—to dominate with them is legendary. Craig is legendary. There have been others, mostly unknown and unheralded, but leaving an imprint on motorcycle history that’s unparalleled and impressive. On our side of the ocean, it amounts to a handful of men over the entire spectrum of American racing. Ottaway heads the list. Not just for the eight-valve either! But the eight-valve, its variations and mysteries, is what we’re here for. So…

Patently Superior

Hereabouts on these pages is a line drawing you might have glanced at a time or two before. Please look at it closely! (I’ll wait here.) It’s the patent drawing for design fundamentals of a V-Twin racing engine, applied for in 1919 and awarded to Bill Harley in 1923. These “fundamentals” amount to the method of valve actuation and the layout of the four valves per cylinder head, as well as details peculiar to the entire concept. Damn near all of it, at some point in the history of the actual machines, tampered with, altered, changed, ignored and in short, different from the drawing! Hey, you’ve got to expect that with a racer… right?

There are also pictures (mostly not that great since quality photos of these eight-valve machines and engines are almost as rare as the bikes), which hopefully reveal just how different the details could get! All in a day’s work at the time… but a real riddle for historians. I mean these things are so rare in the first place, that coupled with all the tweaks and tricks that racers tried on ’em in the day it’s hard to say that any of the handful of survivors are “original.” Not to mention the stuff the factory did (or might have done), courtesy of Ottaway and individual rider preference. Add the fact that factory records are nonexistent, so no real proof exists on the subject one way or another. What are we left with? Mostly, published race results from the time, some photos, the patent and, of course, a precious few machines. The patent is ground zero, race results affirm H-D’s winning ways historically, but it’s the photos, then and now, that are the intriguing part of the legacy. Only those… and the machines that exist today, (including “repros” for the well-heeled) forming tangible, all-too-scarce links to that legacy of “Aces on Eights.”

1 Here we have the illustration for U.S. patent #1,472,068 for an engine. Nowhere in the text does this patent refer to racing, or anything related to high performance. Instead it states the purpose of the patent is to: “insure uniform correlated expansion of the co-operating members of the mechanism. A further objective is to insure regular operation of the engine irrespective of the temperature at which the engine is operated. A further objective is to provide a motor in which the errors of timing of the mechanism due to expansion of the parts are eliminated. A further objective is to provide a lightweight motor.” WTF? Some of this language is probably code for early, extensive employment of aluminum alloy in its construction. But… c’mon!

First of all we know this engine was developed in 1915. (Some say a “prototype” single-cylinder version was even raced late that year.) The complete V-Twin racer was revealed and successfully raced in 1916. Yet, it was cataloged as the Model 17. The war was over in 1918. The patent was applied for in 1919 and granted in 1923. By 1921, the design had changed significantly and at the end of the year… H-D quit racing! If all this isn’t enough to make you curious about what was really going on, the patent refers to a pushrod “housing” (reference #13 in the drawing) intended to “heat” air surrounding the thus-enclosed pushrods, again in the interests of uniform expansion. Yet, these housings weren’t used! No surviving OHV race bike of the period, let alone an eight-valve, has them. (Pushrod tubes, as we know them, were added to street machines several years earlier. But that’s not the same thing, since tubes were for—ah— “oil control.”) And why bother with a patent in the first place, for a machine that was never made available to the public, was only manufactured in miniscule quantities and disappeared from the scene before the patent was even granted?

So, to speculate, could it be that the eight-valve engine was a “lab rat” for a potential street version that never happened? The intent of the patent notwithstanding, upsets of war, the post-war negation of Indian as a threat, and the subsequent economic crash in 1921 might have conspired to keep us from an advanced-spec V-Twin engine for generations!


So here’s your basic 1916 Model 17. Note the engine mounting plates, manual oil pump, the four short head pipes, the configuration of the cam cover, the pushrods and brackets for the rocker arms? This is the racer Harley only used for team riders before World War I. Except for the front forks! (In our last installment, you might have noticed that both the Harley ad and the over-restored 1916 bike pictured feature “girder” forks… not the springer.) Just consider this our first example of modifications to these race bikes.


This 1918 model, if accurately dated, has already been subject to a wad of detailed differences and more than a few major ones. Differences that are easy to spot include the lack of head pipes and the oval-shaped exhaust ports, cam cover configuration, the pushrod bracket redesign, and the fascinating contraption wrapped around the bottom of the front cylinder… which connects to what appears to be an external automatic pump on the cam cover that, in turn, has a fat tube running down to it from the top of the engine. (Fact is, you could specify a manual pump, automatic pump… or both!) The cam lifter (compression release to us) in front is still there, but operated by rod linkages, not a cable. The list goes on, but as you observe, you begin to realize there just might be more differences than similarities in each and every one of these engines. Not to mentions chassis! Sort of a tailor-made no-two-alike deal. Of course it could also be after all these years, that the heads on this one aren’t the same ones it left the factory with, couldn’t it?


This one is a 1923 model… although another eight-valve mystery might be, “How the hell would you know for sure?” Is all this dating business based on when it was built, when it was sold, when it was raced… or what? Harley did ultimately sell some — supposedly only to “qualified” buyers in overseas markets — after they quit using them in the U.S. But did they build a bunch of bottom ends and add the current version of the heads at the point of sale… perhaps years later? Or maybe build the whole thing with what was on hand once ordered… a little of both… or what?

What’s more certain is this particular eight-valve has what should be correct-version heads for the year, and an equally correct version of racing carburetor. But hold on! What’s the deal with the pushrods? A close look at the “tappet block” area sure seems to reveal a total reversal of intake and exhaust cam lobes inside the case… doesn’t it? Jeez! Could that mean one of these twin cam bottom ends is indirect-acting and the other direct-acting?


While you’re pondering that notion, what about this? What we have here is a 1927 eight-valve (recently unearthed in New Zealand, attached to a sidecar) which sold at auction for an astounding $600,000-plus! As dirty as it appears, this one is pretty much the “cleanest” so far and is likely to be representative of the last of them. Look once more at the tappet deck on the twin cam case. Is it me, or are the pushrod bottoms perched side by side, rather than staggered as with the other examples? They run nearly parallel to the rocker arms as well. Those heads are most likely the final iteration (big-valve, oval-ported) just to sweeten the deal. And how about that left side-facing carb… eh? Although you can’t see it in these pictures, the spark plug location in eight-valve heads “migrated” higher and higher in each redesign, until they arrived at the “modern” location above and between the valves. For a sharp contrast (as pictured in part one of this treatise), look at plug location in the Indian 8-V racers! These kinds of subtle variations and refinements in the Model 17 are all the more reason they had such a horsepower advantage.


Finally, and strangely… it’s another 1923 model… with a twist (or two)! This “Statnekov” 8V was acquired from Boozefighter legend John Cameron who bought it from the estate of English racing genius Freddy Dixon. Statnekov finally sold the bike to Harley to put in their museum. Yes, when you see the eight-valve racer that H-D now owns, it’s this one. Sounds simple. Yeah, right! To begin, it’s an early single-cam “banjo” case engine! Were they really still building the first iteration engine in 1923? The heads appear to be the “middle” version of three types used… namely the “kidney” port type, available, by the way, with or without exhaust headers. Dixon was a racer’s racer and to me, it seems a virtual certainty this thing had/has been messed with so much for so long that dating it to 1923 is a dart board guess.


This is partly why! Mr. Dixon himself, standing behind one iteration of his 8V race bike… the one in the Museum! Except, like Washington’s cherry tree hatchet, already, the head’s been replaced once and the handle twice! I mean, look at it! The non-Harley cradle frame has a tube looping under the engine, the tank has cutouts, there’s a hand shifter for the non-Harley two-speed gearbox, and that friction damper on the fork was never a Harley part! There’s rear stand and modified rockers on the springer… shall I go on?


This is partly why! Mr. Dixon himself, standing behind one iteration of his 8V race bike… the one in the Museum! Except, like Washington’s cherry tree hatchet, already, the head’s been replaced once and the handle twice! I mean, look at it! The non-Harley cradle frame has a tube looping under the engine, the tank has cutouts, there’s a hand shifter for the non-Harley two-speed gearbox, and that friction damper on the fork was never a Harley part! There’s rear stand and modified rockers on the springer… shall I go on?


OK… I will! Here’s the same bike, same famous rider/tuner… different date… probably a couple years earlier. Looks a little more “factory”… no tranny, no shifter, no cradle frame. Except for some details, the only really unusual item that meets the eye is that burlap bag-looking thing where the carb should be. Legend has it that there might just be two British carbs mounted under there. The factory never did it… but Dixon probably did.
The point, put in simpler terms, is this bike was/is a race bike. A damn fast, highly competitive and (for its time) technologically superior racing machine for over a decade. All race bikes get tweaked and tricked out from frame, to fork, to engine and every single thing in between, in order to gain a competitive edge, and all eight-valve Model 17s are race bikes. Pedigree is secondary to performance in racing. You’ll never see a “stock” one… even in a museum. I’m just glad these bikes survive (in whatever form) to show us how it was done… and done well… back in the day!


  1. Hi Kip,
    From a Gearhead. Got Feb 17 yesterday.
    Shovels, YEA. Twin Factory Belt YEA. Too bad the primary one was less than spectacular in dependability. Of course the after market had been making their better, but not perfect version of it for quite awhile by then, with the factory try probably motivated from those. ONlY two model years, Sturgis only, WRONG. I do have in my possession a friends 1982 last Shovel Sturgis, which still is factory primary & secondary belt drive, that one makes 3 years 1980 – 1982 factory twin belt, Sturgis model. Ever seen a 1983 FXDWG Disc Glide? my absolute always wanted couldn’t afford to get one. I was unable to get the one and only FXDWG sold at the local dealer, so I ended up with a regular 1983 FXWG as I didn’t have the time money etc to have him try and get another one. Very limited production model, one of, if not the first of “Willie G Specials”, [yes the boattail Super Glide was a W G creation, but I don’t think they start calling them W G specials then.] unusual finish, Maroon, & Maybe Dark Purple, and VERY limited chrome and polish Blacked out instead etc. If I remember right a unless you were a special dealer you got ONE and only one. That makes 4 years 2 models for twin belt. I won’t say I’m 100% correct. But my memory says I’m close. In this twin belt discussion I am referring only to the 4 speed swing arm frame models, 1958 – 1984, & maybe 1985, 1986 re engineered to accept the EVO. Yes I know about the unpopular 1980 FLT Tour Glide, [Another wish I had one bike, Now, Not then.] twin chains, 5 speed rubber mount, which they still had in the EVO early years. Also the 1982 FXR series. The factory didn’t rear belt all big twins until at least 1986 if my memory is correct. I know where one of the above FXDWGs was and probably still is. Unfortunately the friend of mine, [Who’s only forgiveness for this is that he didn’t get it until it had enough modifications to make it just another Wide Glide to those who didn’t know, and passed what is left of it on to his brother, much more non stock.], didn’t stop playing with it until it isn’t one any more, except by VIN, if HD keep good enough records, and they weren’t lost, destroyed ETC. He wasn’t the first modifier of it etc. This is almost all memory. The 1983 Shovels only got rear belt on LIMITED models. Other than the FXDWG & maybe Shriner baggers, I don’t remember yes/no on the rest, except my 83 FXWG was all chain. Didn’t look in detail or remember about the other 1983s. The 1984 Low Riders [& maybe the FXWGs] had rear belt, front chain. Another one my memory is welded to, I lucked into a badly treated 84 Low Rider reasonable and rebuilt it, twin chain, rear off an 83 FXWG, plus other alterations, minor, and resold it.
    It’s rear belt is on my 83 FXWG, with an idler equipped after market Primary belt. A very fun Bike in it’s day. Geared a bit tall but fun. Some other Special models may have got the factory rear belt, in 1983. The 1984s I can’t remember accurately. Back then I still LOVED playing with stuff. Time etc has it’s effects, doesn’t it. That 83 remains my favorite one.
    That turned into a more lengthy ramble than it was going to be, happens when I stir my memory.

    On to other important things, sort of. I’ve enjoyed your TP MM column as long as I’ve been reading it. The Name that Time Machine is less interesting some times and more others, but always Fun. This isn’t the first time I’ve disagreed with you, but this one I’m more convinced of, and do this sort of thing more than in the past. The last was Shovel related long before Time Machine, but a Time Machine type piece in the MM column, Which I remember thinking then wrong, sort of but wasn’t important enough then to say so. [E Mail wasn’t as easy then either.]
    Keep up the good doings enjoy, even when you are some right some wrong, I always enjoy it and frequently learn things. That’s what was attractive about your writings since I started following them.


Please enter your comment!
Please enter your name here