The Destructive Indianapolis Track Narrative of 1919

Unearthing the rough rides of the brilliant Ballots

Arguably the most sophisticated and fastest racing cars on earth in 1919 were the four racing Ballots specially made for that year’s Indianapolis 500. Even before setting a new lap record of 104.7 mph during qualifying, the untried team were the prohibitive favorites to win. That their efforts produced disappointment, two crashes, a fourth-place and a tenth-place finish, is well known.  In Miles C. Collier’s forthcoming book, The Archaeological Automobile, he presents three heretofore unknown but important stories about the Ballots at Indianapolis, the “Cooling” and “Handling” narratives, that were unearthed through archaeological analysis; the third account, “The Destructive Track,” follows below.

Contemporary newspaper reports of Indianapolis in May 1919 make it clear that the track was unusually rough. It had received no maintenance during the two years of inactivity during World War I. A particularly dramatic newspaper account describes one of favored-to-win Rene Thomas’s hot practice laps during an early test session. Entering a corner at the limit of control, Thomas hit an unseen diagonal wave in the brick track that launched his Ballot off the pavement. The car began to rotate in the air in such an alarming manner that a major crash appeared inevitable. Only Thomas’s brilliant driving allowed him to regain control. So rough was the track that one reporter noted that the drivers had to memorize every bump and subsidence all the way round the circuit in order to ensure their safety, to say nothing of cutting a quick lap. Experienced commentators, who had probably been talking to drivers, said that the Indy cars of 1919 were too fast for the condition of the racing surface.

So, how rough was the track really? What effect did this brutal surface have on the cars and the drivers? What, if anything, did the teams do to combat it? These are the types of questions archaeologists might ask, and about which historians can supply no answer because such information was never recorded. Yes, from contemporary accounts, historians know about the damage to cars during the race that can most appropriately be ascribed to the bumpy surface. Eyewitness reports describe Ralph DePalma’s race-leading V-12 Packard retiring with a broken wheel bearing; collapsed wheels were common, fatal roll-over crashes occurred, the first since the inaugural Indy 500 in 1911. Pit stops for suspension repairs were legion. But can we find non-text-based evidence – evidence from physical matter – that shows, not tells, what it was like in real human terms?

Finding archaeological data is one thing, but the interpretation of such data requires a specific tool. And “archaeological imagination” is that instrument. It allows us to plausibly reconstruct the how and why of a discovered relic’s design, making, and use. It allows us to uncover lost events and narratives, to create reconstructions, conjectures, and hypotheticals from ancient evidence. And this proved to be the case with our Ballot.

Fortunately for our restoration team, we had additional archaeological matter to consult, a small collection of photographs from 1919 constituting the known photographic record of the Ballot team at Indy. Photographs are every bit as much archaeological relics as are pottery, bones, or the Ballot itself, because historical photographs are the visual records of instants of reality from the past. By consulting the surviving 100-year-old images and employing our archaeological imagination on them, we quickly established that, while the Ballots were in a state of continuous change during their time at Indianapolis, the team was primarily concerned with three common issues that race teams face to this day: reliability, cooling, and handling. In recent work with our Ballot, we discovered specific evidence of their struggle to ensure 500 miles worth of reliability on Memorial Day given the track’s bumpy surface and the toll it was exacting from the cars during the run up to the race.

A photograph taken at the Paris factory shows the side view of a brand-new Ballot, taken perhaps even before its first road test in March.

We see an immaculate and beautiful machine that displays all the elegance inherent in its superior design. By contrast, photographs taken a few days before the Memorial Day race show the same car after three weeks of testing and tuning in distinctly “used” and roughly modified condition.

Gone is the immaculate elegance, now present are a host of rough and ready modifications: a hand grip for the riding mechanic behind the driver’s right shoulder, a driver’s wind deflector, a much more substantial radiator cap with condensing coil and drain hose, additional taping on the rear springs, prominently installed large, “scissor” type friction shocks on all four wheels, and additional cockpit padding.

The auxiliary shocks speak directly to the track conditions. The elegant French front drum and annular rear units were insufficient for conditions. Hence Rene Thomas’s scary moment that occurred shortly after the team’s arrival at Indy early that month.

Close examination of the body of Thomas’s Ballot from photographs taken just before race day shows additional fasteners securing the aluminum body skins into the wooden ledgers that through-bolt to the chassis rails.

Cars of the period had relatively flexible chassis and relatively stiff suspensions. Consequently, twisting and racking of the chassis during operation was a normal phenomenon. Indeed, to isolate the chassis flex from the engine, a three-point, “floating” subframe was part of the Ballot design. The subframe could pivot as the chassis flexed, preventing twisting torque from being experienced by the powerplant.


From the additional fasteners holding the body to the chassis, we can infer that so extreme was the chassis flex caused by the rough track, that the team feared that the original body bolts would literally tear out of the thin aluminum skin. The team’s solution was the addition of a lower row of, not bolts and anchors as used in the original design, but screws driven through the body into the ledgers.

From inspection of the seats shown in photographs, it is notable that Ballot seat upholstery was made to a very loose, rather sloppy looking design. The “squishy” nature of the softly padded seats required the occupants to “wiggle” themselves into the grip of the padding, thereby achieving some degree of restraint in an age before seatbelts were used for that purpose.

Despite the grippy seats, the riding mechanics evidently discovered that they needed something more, hence the steel grab handle attached to the body behind the driver’s right shoulder.

An additional leather handle was also attached to the left-side cockpit coaming for the mechanic’s security in the event he had to attend to the firewall mounted lubricators.

The driver of course had the steering wheel for security, but that, too, presented problems. Remember that Indianapolis consists of four left-hand turns. Inertia would press the drivers against the right side of the cockpit. To resist the cornering force, drivers would use their right knees to brace against the body side so that they could maintain a delicate touch on the wheel. And this was just as well, because these early racing cars were unequipped with anything like a modern instrument panel that could support the steering shaft with a bracket close to the wheel. Instead, steering columns were secured by one or more bracing struts originating at the firewall. A photograph of Thomas’s car in the pits later in the race shows what is apparently a leather belt wrapped around the steering column at the point where the two steering column bracing struts attach.

The right or driver-side strut must have fractured at the steering column attachment point. Arguably, the merciless pounding dealt by the track, abetted by Thomas’s hanging on the wheel, proved too much for the right-side strut and a hastily improvised track-side repair with a leather belt was the result.

But the team’s problems had not stopped there. Paul Bablot’s number 33 Ballot crashed out of the race with a broken rear wheel, thankfully with no injury to the occupants.

This rear-view photograph of the wrecked car shorn of its transverse bolster gas tank shows a remarkably crude, unpainted, flat steel strap connecting the two seat backs. This strap is nothing like the high standard of the Ballot cars’ French workmanship, and clearly originated at Indy. We can only surmise that the pounding of the track was so severe that the seats were flexing excessively. The steel strap added badly needed bracing.

If all this is not enough mechanical carnage, a picture of Thomas’s car in the pits during the race itself shows a broken rear axle limiting strap.

These very strong composition canvas straps were installed to prevent the axle from overextending the springs and shocks after large and violent deflections caused by bumps. The car’s left rear strap, which is subject to extra strain from the left-hand turns, has snapped. One can only imagine how many times the big, heavy axle had to slam against the snubber to break it.

Finally, historical records do give us one piece of human color: Ballot team driver Albert Guyot retired early in the race with such damage to his hands from the flaying they received from wheel kick-back over the bumps that he was replaced by reserve driver Jean Chassagne. Chassagne almost immediately experienced a wheel failure–caused crash that ended his effort, fortunately without injury to himself or his riding mechanic.

From historical accounts, we have always known the track was rough in 1919. What those accounts don’t tell us is what the actual experience was like and, further, they are silent as to the feverish struggle that the Ballot team, and presumably the other teams as well, experienced in their attempts to deal with the extraordinary punishment meted out by the track, both before and during the race. Through archaeological evidence, and through using archaeological imagination to connect clues from the past with a web of narrative based on our understanding of historical racing cars and motor racing practices, the problems the Ballot team encountered, and the solutions they found, are now made current, understandable, and very real.

Miles C. Collier’s book, The Archaeological Automobile, will be published in the fall. To learn more and to preorder, go to