Deep in the mountains of the Pacific Northwest, sometime in the 1890s, a locomotive was climbing a grade that no railroad in the country would have permitted on its main line. The track beneath it was not ballasted with gravel. It was not even perfectly straight. It had been laid on bare ground only a few weeks earlier, and would be torn up and moved again within months.

It curved around tree stumps. It dipped between wooden ties, and up this rough, temporary, impossibly steep grade, a machine weighing as much as a loaded freight car was hauling itself and a string of log cars behind it, one slow yard at a time. No ordinary locomotive could do this. An ordinary locomotive set on this slope would spin its wheels, lose its grip, and slide backward down the mountain.

Every railroad engineer in America knew that, and that was precisely why the timber here had been considered unreachable until a lumberman from Michigan built an engine that broke the rule. A conventional steam locomotive is a delicate thing about where it can go. It has a set of large driving wheels, and the engine turns only those wheels, through rods. Every bit of the locomotive’s pulling power passes through those few driven wheels, and all the traction those wheels have comes from the weight resting directly on top of them.

Put that machine on a steep grade or a rough, uneven track, and two things go wrong at once. The driving wheels begin to slip because there is not enough weight on them to hold. And the long, rigid wheelbase does not want to follow a track that twists and drops. So main lines were built flat.

A one percent grade, one foot of rise for every hundred feet of track, was normal. Two percent was considered steep, the kind that made engineers nervous and required helper locomotives. Anything beyond that simply was not built, because a conventional locomotive could not climb it reliably. Now consider where the timber was.

The largest and most valuable forests in America were not conveniently located on flat ground beside a river. They were on mountainsides above grades, in country that rose and fell and twisted. For most of the nineteenth century, the only way to get timber out of such country was with animals and gravity: teams of oxen on greased skid roads, timber chutes, and flumes that angled down across the mountain. All those methods worked, but each had a ceiling.

They could only reach so high, and they could only bring timber down, without the ability to move operations upward. What the lumbermen needed was a railroad that could climb. By traditional railroad standards, that was a contradiction in terms. The man who solved the problem was not a locomotive engineer.

He was a lumberman. His name was Ephraim Shay. He was born in July 1839 in Ohio. He tried many things in his life.

He taught school. He served in the Civil War. He ran a store. But by the 1860s he was working in the timber trade in Michigan, near a town called Cadillac, wrestling with the same problem described above.

He was trying to get logs to his mill, and the available roads, which in Michigan meant hauling on sleds over winter snow, were slow, seasonal, and maddening. He wanted something that could work year-round and reach where the timber was. Beginning around 1872, Shay began to build it. What he built looked like no locomotive anyone had ever seen, because he discarded the single assumption that makes a locomotive a locomotive.

It is worth pausing over the nature of the man, because it explains the machine he made. Shay was not a trained engineer working from theory. He was a practical craftsman of the kind the nineteenth century produced in abundance, a man who looked at a problem in front of him and built his way through it. Later in life he designed and lived in an all-steel house shaped like a hexagon.

He also built and operated a waterworks for his town. He was exactly the sort of person who looks at a locomotive, decides the standard design is wrong for the job at hand, and simply builds a different one. The Shay locomotive did not come from the great railroad works of the East. It came out of a Michigan logging operation, made by a man who needed to move his logs and was not prepared to accept that it was impossible.

Here is the part that is hard to believe. Shay’s idea was to stop moving just a few wheels, and instead drive every wheel under the machine. In a conventional locomotive, the cylinders drive rods that turn the main driving wheels. That is all.

In Shay’s design, the cylinders were mounted vertically on the side of the boiler, turning a crankshaft that ran the entire length of the locomotive. From that crankshaft, a line of shafts and bevel gears connected to every axle on every set of wheels, including the wheels under the cars carrying the water and fuel. Every wheel on the machine was a driving wheel. Think about what that changes.

On an ordinary locomotive, only the weight over the driving wheels gives traction. On a Shay, the entire weight of the machine, up to eighty tons in the largest models, rests on wheels that all pull at once. There is no dead weight being dragged. Every ton of the locomotive is a ton of traction.

And because the power is distributed over many small wheels turning slowly through gearing, rather than a few large wheels turning fast, the machine could climb patiently with a pulling power no rod engine could match. It was not fast. That was the trade-off. A Shay moving at more than ten or fifteen miles an hour was moving at a lively clip.

But speed was never the point. The point was its ability to climb grades above five percent, grades several times steeper than anything a main line would dare, and in some logging lines steeper still. It climbed them slowly and relentlessly, without slipping, hauling a load far greater than its size suggested. To put a real number on what that meant, look at one line still operating today.

At Cass, West Virginia, a logging railroad was built beginning in 1901 to climb Cheat Mountain and reach the timber at its summit. The line rises about 2,400 feet over eleven miles of track. In places, the grade reaches eleven percent, eleven feet of rise for every hundred feet of track. For comparison, remember that main-line railroads begin having trouble at two percent.

This was more than five times that, and Shay locomotives climbed it hauling loaded log cars. One of the machines that did it, an eighty-ton Shay of the C class, has been working on that mountain for nearly a century. Cass exists today as a preserved railroad precisely because what happened there was not supposed to be possible by ordinary railroad standards. There was a second trick used to climb a mountain that steep, a trick that looks almost like cheating.

When even a Shay could not climb straight up, the loggers built the track as a switchback. The train climbs one section of track to a dead end with a switch, stops, throws the switch, then backs up the next section in the opposite direction. It climbs all the time, back and forth, gaining elevation with each leg. The entire train reverses direction again and again to work its way up the side of a mountain where there was no room for a gentler slope.

Cass has two of them on the way to the summit. The crew working a switchback line would back eighty tons of locomotive and load up a temporary track, and they did it as daily routine. The same gearing that enabled the climb enabled the far more terrifying part: coming back down. Because the wheels were connected to the engine’s gears, the engine could brake with them.

Descending a loaded grade, the machine could use its own gears as a brake against gravity, critical because the alternative, losing control of a heavy load on the way down, was what killed many men in this work. The design had a price. Look at a Shay and it will look tilted, almost wrong. The boiler is offset to one side.

That is not a mistake. The vertical cylinders and heavy gearing hung entirely on the right side of the machine, so the boiler had to be shifted left to balance the weight. It looked strange, but it worked. Once loggers saw one running, they wanted one.

In 1880, the first of these locomotives was shipped to a customer in Michigan. In 1881, Shay began taking out patents. He granted a manufacturing license to an Ohio plant that grew into the Lima Locomotive Company, and Lima began producing them. Hundreds were built in the first decade, and eventually more than two thousand Shay locomotives of various sizes were sent to logging and mining railroads across the Americas, and as far away as Australia and East Asia.

A lumberman’s answer to Michigan snow had become a machine used on three continents. Shay’s idea was good enough that others built their own versions. Within a few years, two serious competitors appeared, and together the three defined an entire class of machines that existed only in this kind of work. There was the Climax, which angled its cylinders and ran its drive shaft down the center of the locomotive.

There was the Heisler, with a cylinder on each side angled inward in a V, generally considered the fastest of the three. Each had its advocates, and each had real advantages. The Climax and Heisler ran their drive shafts along the centerline, which some argued gave them smoother motion on rough track. But the Shay remained the most popular.

The reason it stayed the most popular tells you something about who was actually using these machines. The Shay carried its machinery on the outside. All those shafts, gears, and cylinders were openly exposed on the side of the boiler, where a crew at a remote logging camp, with no proper roundhouse, no machine shop, no factory nearby, could see them, reach them, and repair them. A locomotive that a logging crew could keep running themselves in the woods was worth more than another that was a little faster but required taking it somewhere to be fixed.

The men who bought these machines were not railroaders. They were lumbermen, and they bought the machine a lumberman could keep working. Now for the work itself, because a machine that could climb a mountain still needed men to run it, and running it was unsafe. Start with the track.

A logging railroad was not a permanent thing. It was laid to reach a particular stand of timber, used until the wood ran out, then torn up and laid somewhere else. This meant the track these eighty-ton machines ran on was terrible by any ordinary railroad standard. Light rail, sometimes just timber topped with iron straps.

Ties laid on bare ground. Sharp curves around obstacles, grades that changed suddenly, and wooden trestles thrown up quickly from the nearest trees to cross the nearest ravine. The Shay was designed to tolerate exactly this kind of track. Its wheels could swivel and follow a line that would have derailed a rigid locomotive.

But tolerating bad track and staying safe on it are two different things. There were a great many of these railroads. This was not a handful of unique lines. Across America’s logging regions, tens of thousands of miles of temporary track were laid over the decades to reach the timber.

A whole hidden railroad network, built to be laid, used, and pulled up, following the cutting ever deeper into country no permanent railroad would ever enter. Every mile of it needed crews to lay it, crews to run it, and crews to tear it up again. And all of that work was done on grades and curves that a main-line engineer would have refused to approve. Then add the load and the grade.

A Shay coming down the mountain with a string of loaded log cars behind it was handling enormous weight on a slope and a temporary track, using gears and hand brakes to hold it all back. Think specifically about the descent. Going up, the danger is that the locomotive stops. Coming down, the danger is that it does not stop.

Here gravity takes hold of eighty tons of locomotive and a train of loaded log cars and begins pulling everything down the mountain faster than gears and brakes can stop it. The engineer runs the engine against the grade to hold the train back. The brakeman rides the cars, ready to set the hand brakes. And everyone aboard trusts light, temporary rails laid on bare ground weeks earlier to stay under the wheels through every curve.

If anything failed, if a brake would not hold, if a coupling broke, if a rail gave way on a curve, the train was going down fast, and the men aboard faced that outcome. A logging train running out of control was not a minor accident. It meant the locomotive and cars leaving the track on a mountainside. The men who rode these trains lived in a world with none of the protections now taken for granted.

No signals. No brakes reliable by modern standards. No rulebook worth trusting in the early years. Crews learned the nature of the grades by working them.

They learned which curves demanded caution. They learned how much load each section of track could bear. Those who learned the wrong way did not always get a second chance. There is no accurate count of how many men died operating logging railroads.

The reason is the same one that recurs throughout the history of this work. The operations were in remote places. The crews were transient and unstable. Companies kept records of board feet of timber and locomotive repairs.

They did not keep records of men. A brakeman killed on a mountainside grade at an obscure logging camp in 1894 ended up in a grave. If he was lucky, his name appeared in a line in a local newspaper. So when it is said that this work was dangerous, understand that the absence of numbers is not evidence of safety.

It is evidence that no one was counting the dead. As always happens, the era of the geared locomotives ended not because the machine failed, but because something better appeared. The Shay had a weakness, the other face of its strength. It was slow and mechanically complex.

All those shafts and gears demanded constant maintenance. As long as the job was dragging heavy loads slowly up impossible grades, that did not matter, because there was no alternative that could do the job. But once other technology caught up, the Shay’s disadvantages began to tell. What replaced it was the internal combustion engine.

A logging truck could go where no railroad, not even a Shay, could reach, over a dirt road cut in a single afternoon, with no rails to lay or tear up. And when the diesel locomotive arrived, it could do main-line work faster and cheaper than any geared steam engine. One by one, the dedicated logging railroads were pulled up for the last time and never relaid. The Lima Locomotive Company stopped building Shays.

Those machines were scrapped or left to rust on spurs, while a few were preserved on tourist railroads, where you can still hear one climbing a grade with that distinctive accelerating exhaust beat that makes it sound as if it is moving far faster than it actually is. But for roughly half a century, the Shay accomplished something that by every standard had been considered impossible. It put a railroad in places railroads could not go. It allowed loggers to drive eighty-ton locomotives and their loads up grades that main-line engineers would not have dared attempt, over rails laid directly on bare soil, to reach timber that had remained untouched on mountainsides precisely because no one could get it out.

And the man who made it possible was not a great railroad engineer or a famous inventor. He was just a Michigan lumberman who grew tired of waiting for snow, looked at the one rule everyone said a locomotive had to follow, that it could only drive a few of its wheels, and decided to drive them all instead. The mountains did not get smaller.

The machine just got smart enough to climb them.