In the summer of 1925, a survey crew from the Bureau of Public Roads stood at the base of a formation called the Garden Wall and studied a problem that had defeated every road engineer who had examined it before them. The Garden Wall is not a hillside, not a steep slope, but a rocky projection, a thin blade of stone formed where two glaciers carved the mountain from opposite sides simultaneously over ten thousand years, leaving behind a wall so narrow at its crest that a person could stand on top and see both sides of the Continental Divide at once. On the eastern face, the wall drops nearly two thousand feet to a lake called Saint Mary. On the western face, it descends into the MacDonald Creek valley.
Both faces are vertical in many places. The plan called for a road to cross this wall, not over it, but through it, carved into its face. A rocky shelf would be cut at mid-height, wide enough for two cars to pass, clinging to the cliff for twelve miles at elevations between four thousand and sixty-six hundred feet, passing through sections where the mountain rises another thousand feet of unstable rock above the road, with the void below falling away so steeply that a man falling from the outer edge would not touch the slope before reaching the bottom. The engineers who surveyed the line understood the problem.
Their surveys let them determine direction and grade, and tell the construction crews what height and width to cut. But the surveys did not tell them where the rock would cooperate and where it would not. The Garden Wall was not uniform. It was ancient sedimentary rock, Precambrian limestone and argillite, accumulated over a billion years, then thrust upward by the Lewis Overthrust, one of the largest geological events in North American history, which moved an entire plate of ancient seafloor more than eighty kilometers eastward over younger rock beneath it.
The result was a mountain range built of rock far older than the terrain it sat upon, fractured in patterns no surface survey could fully predict, subject to conditions no construction crew working in a desert canyon or a river valley had ever faced. Snow. The Garden Wall receives more than 250 centimeters of snow annually. The construction season at this elevation and latitude was five months at best, from late June when the snowpack melted enough to reach the site until mid-October when it closed again.
In some years, it was four months. In some years, less. The crews did not work through the winter. They could not.
The threat of avalanches above five thousand feet in the northern Rockies in winter is not a manageable risk but a condition that ends human existence on a mountainside. The crews worked five months, descended before the snow closed the roads to the mountain, and returned the next summer to find the winter’s effects on what they had built. The effect of winter on what they built was the problem they solved every spring before they could build anything new. To understand why anyone would attempt a road across the Garden Wall, you must understand what a national park meant in 1924 to the people who ran the railroads.
Glacier National Park was established in 1910. The Great Northern Railway had been running tourist trains to the park’s western entrance since 1892, when James Hill’s line reached the Flathead Valley. Hill understood something about mountains that the automobile age would confirm. If you can show people the interior of a mountain range, they will pay to see it.
The problem was that the interior of Glacier was not accessible. The park’s most spectacular regions, the backbone of the Continental Divide, the glaciers, the passes, the peaks from which the park took its name, lay behind the Garden Wall, reachable only on horseback over trails that took many days and required guides, horses, and camping equipment, limiting the experience to visitors who had abundant time and money. Automobiles were changing what Americans expected from national parks. By the early 1920s, the number of cars in the United States had risen from a few thousand to nearly ten million.
People who could not afford a week on horseback could now spend a weekend in a car. Parks with roads drew visitors; parks without roads were silent. The National Park Service, founded in 1916, understood this calculation. Its first director, Stephen Mather, was a businessman who made a fortune in borax mining before taking the federal post.
He believed the parks would be saved by their visitors, and that those visitors would come by road. The Going-to-the-Sun Road was his project as much as anyone’s. A road designed to cross the Continental Divide at Logan Pass, connecting the western entrance at West Glacier to the eastern entrance at Saint Mary, allowing a drive from one side of Glacier National Park to the other in a single day. That required cutting a rock shelf into the face of the Garden Wall for twelve miles at a medium elevation.
The Bureau of Public Roads surveyed the line in the early 1920s. The survey was done on foot by men who walked the cliff face carrying compasses and altimeters, recording bearings while standing on trackless rock. When construction began in earnest in 1925, the crews had only a line on paper, and nothing else on the mountain. The men who worked the Garden Wall in the summer of 1925 were drawn from the skilled labor pool of the northern Rockies, which in that decade meant a specific set of communities.
There were men from the farming towns of the Flathead Valley, who spent their first winters on their homesteads after the land sessions of 1910 and found that high-altitude farming on a short season did not support their needs. There were men from the mining camps of Butte and Anaconda, skilled in rock work and powder, undisturbed by tight spaces or the sound of blasts. There were also men from the Blackfeet tribe, whose reservation borders the park on its eastern edge, whose ancestors had used the passes through this mountain range for centuries before the park was declared and before the road was even conceived. And there were Italian stonemasons.
The Bureau of Public Roads brought in skilled masons for the project’s masonry work, from retaining walls and drainage channels to guardrails and structures supporting the road base on the cliff face, preventing the outer edge from collapsing under load. Many of these men came from communities in Montana, Colorado, and Wyoming, where Italian immigrant workers had been doing stonework since the railroad era. They were the second generation of that immigration. Their fathers had set dynamite on mountain railroad lines.
They themselves carved and placed stone on mountain roads. The difference was in the precision required. A retaining wall carrying passenger cars had to meet standards a railroad fill did not. It had to look right and be durable.
Wages on Garden Wall construction varied by skill and task. Common laborers earned $3. 50 a day in the mid-1920s. Skilled drilling crews earned between $4.
50 and $5. Powder men earned slightly more for the specialized knowledge required. The Italian masons, brought in on contracts rather than day labor, negotiated piece rates for specific structures. In today’s value, $3.
50 in 1925 is equivalent to about $62. For a ten-hour workday on a rock cliff at 5,500 feet in the northern Rockies, carrying tools over trails the road had not yet reached, under a schedule that closed the mountain in October regardless of progress, the pay was $62. There was no overtime, no hazard pay, no formal accident insurance. The Bureau of Public Roads kept injury and death records of the project with the same clipped directness it used to account for material costs.
The records read: injured workman, date, cause. Deceased workman, date, cause. Names appear once. You need to understand the specific physical problem of the Garden Wall section before you can understand how it was solved, because the solution only makes sense in light of the reality of the work they were doing.
The average slope of the wall where the road crosses it, between the Loop and Logan Pass, is between 60 and 80 degrees on its western face. A 60-degree slope is steep enough that a person without trail experience could not walk it without using their hands. An 80-degree slope is practically a cliff. The road base had to be cut into this wall at a constant elevation, meaning that on the steepest sections the men were not excavating a slope, they were cutting a horizontal shelf into a near-vertical wall.
To do that, they had to reach the wall. There was no trail above the road line because the road itself was what was being built. Before any blasting on the main road base, the crews had to cut access trails down the slope to the working elevation. These access trails were themselves construction achievements, narrow switchbacks cut by hand through talus and loose rock.
The slope was so steep that equipment and supplies were carried up by mule or by hand, because wheeled vehicles could not manage it. Everything that reached the work face on the Garden Wall arrived on the back of a man or a mule, from steel drill bits and wooden handles to powder boxes, water, food, timber for scaffolding, and rope. All the workers climbed up via the access trails to the site in the morning. They came back down in the evening, because there were no camps on the Garden Wall.
The main camp was at a place called the Loop, a large sharp switchback on the west-side road, where the elevation was lower and the ground level enough to pitch tents and set up cooking areas. From the Loop to the highest work sites on the Garden Wall, the round trip was between four and six miles, with an elevation gain of between 1,500 and 2,000 feet. The men covered this distance twice a day, before and after their shift. This walk was not part of the shift.
It was simply the prerequisite for reaching it. At the work face, the operation involved surveying the line, drilling holes, blasting, removing rubble, and moving the line forward. Drilling was done by two-man teams. The first man pressed a drill bit against the rock surface and rotated it between blows.
The second man struck the drill head with a four-pound hammer at a steady rhythm, two to three blows per second. With each blow, the hole advanced a fraction of an inch. In granite, progress was slow. The rock on the Garden Wall is primarily argillite, a metamorphic stone harder than sandstone but less resistant than pure granite.
An experienced team could advance a drill hole eight to ten inches per hour. A standard blasting round on the Garden Wall required a pattern of holes drilled at specified depths and angles. Usually six to ten holes per round, depending on the width of rock to be removed. The angle of each hole determined the direction the blast would throw the material.
A correctly drilled round on a vertical face would push the shattered rock outward and downward, away from the mountain into the void. A badly drilled round would shatter the rock unpredictably, potentially undermining natural ledges above the blast area or leaving irregular blocks of broken stone that would fall on the crew during the mucking phase. The powder man loaded the holes in sequence and fired the shot. After the dust settled, the mucking crew went to work, breaking up remaining fragments with heavy hammers, then gathering the rubble into wheelbarrows or canvas sacks and hauling it to the outer edge of the road base under construction, where it was dumped over the cliff.
Each blast advanced the road base by two to four feet. To cut twelve miles of rock shelf into the Garden Wall, they repeated this process thousands of times over seven construction seasons. The specific problem that caused the most dangerous work on the entire Going-to-the-Sun Road was not the Garden Wall itself, but a section called the Weeping Wall. The Weeping Wall is a rock cliff on the upper western approach to Logan Pass, where groundwater seeps through the rock and emerges continuously from the face year-round.
Even in winter, when the exposed surface is frozen and the seepage continues behind it, working its way through the argillite, water passes through cracks where accumulated snow and summer rain filter through the mountain and emerge at road level. In summer, water runs down the wall. In spring, the road is ice. The road passes under a rock overhang in this section, blasted to create a clearance of twelve feet above the pavement.
The blast that created this overhang required drilling into a wet face. The presence of water in a drill hole affects how the charge is handled. Drilling wet rock produces a mixture of fine stone dust and water that accumulates in the hole with the drill bit and must be constantly removed to allow the bit to advance. The drill crews at the Weeping Wall wrapped cloth around their hammer hands to absorb the spray from the seeping face.
They worked in wet clothes in an area where the temperature rarely exceeded 50 degrees Fahrenheit, even in July, because the overhang blocked direct sunlight and the wall threw cold groundwater onto anyone standing beneath it. Below the Weeping Wall, the slope drops at about 40 degrees for several hundred feet before falling away into a vertical cliff. A man who slipped from the work face at the Weeping Wall did not stop until the slope ended. The Bureau of Public Roads records show that work on the Weeping Wall section took two full construction seasons.
One season to establish the road base beneath the overhang, and another to complete the blasting of the overhang and build the retaining structure that anchors the upper edge of the roadway against the seeping face. Two full years of wet drilling and bitter cold. There are no names on any sign at the site today. In summer, visitors to the Going-to-the-Sun Road stop at the pullout below the Weeping Wall, walk to the overlook, feel the spray of water falling from the cliff on their faces, and take pictures.
The work that allows them to stand there is invisible from where they stand. Rope work on the Garden Wall followed the same logic as all other cliff construction projects of that era. The most experienced men went down from the edge. On the steepest parts of the Garden Wall, where the angle exceeded a man’s ability to stand without assistance, drill crews worked from rope scaffolding lowered from above or from wooden platforms anchored to the road base under construction.
The wooden platform was assembled from materials brought up by mule over the access trails, fitted together by hand without power tools, and fixed to the rock face with iron pins driven into hand-drilled holes. Part of a platform on the Garden Wall might be eight feet wide and six feet deep, suspended over a void hundreds of feet deep, resting entirely on iron pins driven into the face of a wet cliff by a man hanging from a rope. The man who drilled the anchor holes was not a carpenter or an engineer. He was a rock worker who had spent enough time in the mountains to understand where a driven iron pin would hold in argillite and where it would not.
This knowledge was not written down anywhere. It was in his hands and eyes from years spent in this rock. He drove the anchors. The carpenter built the platform.
The drill crew worked on it. The logic of this arrangement was that if an anchor failed, the man who drove it would not be on the platform when it fell. He would be hanging from a rope beside it, drilling the next hole. Consider this scene with the precision it requires.
You are sitting on a wooden plank suspended by two ropes over a void two thousand feet deep. Your back is against cold argillite, even in July. The midday light reflects off the valley below in a mesmerizing way, though it changes nothing. The men working your belay rope stand forty feet above you on the road under construction.
You communicate with them by pulling on the rope. One pull, lower me. Two pulls, stop. Three pulls, haul me up immediately.
You hold a steel drill in your left hand, its tip against the rock. Your partner, on his own rope, six feet to your right, swings the hammer. The wall is damp from the morning mist that fills the valley below Logan Pass on most July mornings. The drill slips if your angle is not perfect.
You correct it. The hammer swings. You feel the shock travel through your forearm and into your shoulder. You turn the drill slightly.
The hammer swings again. After two hours of this, you will have an eight-inch hole and a bruise running the length of your left forearm from the vibration, which will last two weeks. Then you will come down from the wall and eat cold beans from a metal plate, and return to the rope the next morning. The Blackfeet workers appear in the Bureau of Public Roads construction correspondence as local laborers.
In payroll records, they appear as common laborers, with no further designation. Field supervisor correspondence occasionally notes that certain workers showed good knowledge of mountain terrain, which helped in locating trails in the upper regions. The phrasing is administrative and does not clarify the intended meaning. The meaning is that men whose families had crossed the Continental Divide passes for generations before the park was created were being paid $3.
50 a day to guide Bureau of Public Roads engineers to sites their surveys could not accurately locate from the valley floor. This knowledge was not expensive. It was included in the daily wage. There is a landmark on the Going-to-the-Sun Road that embodies what the masonry crews accomplished and why it matters to this day.
The Triple Arches are part of a stone retaining wall on the western approach to Logan Pass, visible from the road where the roadbed crosses a watercourse descending from the upper Garden Wall. Three stone arches, built without mortar, fitted from argillite and limestone quarried from the surrounding terrain, carry the road over a stream that in spring carries enough water to undermine any poured concrete structure that had not been built with adequate regard for the movement of material beneath it. The arches are built dry, without mortar. Each stone was selected from the valley by an experienced mason who understood the weight it would bear and the geometry of the arch it formed part of.
The keystone of each arch is fitted with such precision that the pressure of the stones on either side holds it in place. Remove the keystone and the arch collapses. Leave it, and the arch grows stronger under pressure, because the pressure compresses the joints and forces the stones to grip each other more tightly. The mason who built the Triple Arches worked from knowledge, not drawings.
The drawings existed; the Bureau of Public Roads provided specifications for all masonry structures on the project, but the drawings showed only dimensions and materials. They did not show which stone to choose from the surrounding terrain. They did not show how to read argillite for fracture lines that might cause a stone to split under pressure after a decade, while appearing sound when installed. They did not show how to level the bearing surfaces of each stone so the contact between stones was flat enough to transfer load without creating a stress point that would crack in the freeze-thaw cycle.
This knowledge the mason carried himself, from wherever he learned it, whether in an Italian mountain village or on a Colorado railroad line or both. It stayed with him. He applied it to a stone wall in Glacier National Park in 1928, and the wall still stands today. The freeze-thaw cycle at 5,000 feet in northern Montana is relentless.
Water seeps into rock cracks in autumn, freezes in November, expanding, pushing the rocks slightly apart. It melts in April, contracting. This repeats every winter. It has continued for nearly a hundred years since the Triple Arches were built.
The arches still stand, and the road still crosses them. Every spring, when the snowplow makes its first pass over the Going-to-the-Sun Road to open it for the season, it crosses the Triple Arches, built by a man whose name appears on no memorial within a hundred miles of where he worked. The Going-to-the-Sun Road was completed in 1932 and officially opened on July 15, 1933. The ceremony was attended by officials from the National Park Service, the Bureau of Public Roads, and the Great Northern Railway.
Stephen Mather, the visionary who had driven the project from its start, died in 1930, two years before the road was completed. He never saw it open. The men who built the road were not present at the opening ceremony; they had moved on to their subsequent projects. Official records put the total construction cost at about $2.
5 million, equivalent to roughly $45 million in today’s prices. The records include materials used, contracts, and completion dates. The names of day laborers appear in payroll records, preserved and available to researchers. But what these records do not show is the nature of each worker’s job; they merely record name, daily wage, and number of days worked.
These records do not distinguish between the man who pushed a wheelbarrow on level ground and the man who drilled anchor holes on a rope over a two-thousand-foot void. The road was designated a National Historic Landmark in 1983, in recognition of the engineering achievement and the aesthetic integration with the surrounding landscape. The official document describes the road as an outstanding example of harmony between built infrastructure and natural terrain, and this description is accurate. The road blends into the landscape as if it were part of it, its curves following the natural lines of the mountain, its stone walls matching the color and texture of the surrounding rock.
This is not coincidence. It is the product of men who chose each stone from the mountain they built in, and placed it with such precision that the joints between stones disappear into the geological formation. The harmony between road and mountain is the product of their labor. The construction of the Going-to-the-Sun Road caused worker deaths; the Bureau of Public Roads records document 11 deaths during the project period.
The causes listed include rockfalls, falls from the ground, and equipment accidents. Eleven is the officially recorded number. This number does not include the men who left the project with injuries that eventually killed them, or the men who inhaled rock dust for six construction seasons and died of lung disease in the 1940s in a town in the Flathead Valley, with no recorded connection to the road they had worked on twenty years before. By 1925, silicosis from rock drilling was understood in a way it had not been in 1879.
The medical knowledge existed, but its industrial application did not. It was known that wet drilling, passing water over the drill head to suppress fine rock dust, reduced silica exposure. It was not standard practice on Garden Wall construction, because water had to be transported to the work site, and the drill crews were already carrying everything they needed up the two-thousand-foot access trail. Adding water supplies to this logistical chain required a decision by someone with authority.
And the men whose lungs would bear the consequences of that decision not being made were not the ones holding that authority. This sentence is what stays with you when reading the Bureau of Public Roads field reports about the Garden Wall construction seasons. The wet drilling decision was a question of logistics and cost, and it was decided on that basis. The argillite dust they drilled entered their lungs because no one with the authority to make that calculation was aware of the result.
Today the Going-to-the-Sun Road is open from mid-June to mid-October, the same season that constrained its construction. The road receives about two million visitors annually. It is managed by the National Park Service as a historic road, meaning modifications must preserve the character of the original construction. When sections of retaining wall collapse, they are repaired with the same material and the same original technique: dry-set argillite, fitted by eye without mortar.
There are still masons doing this work today, who learned it from studying the original walls and from working with men who in turn worked with those who mastered the craft. The chain of knowledge continues. The site has passed through many hands since the men who built the Triple Arches set their last stone and walked back down the access trail in 1928. Those trails are gone; they were temporary routes cut during the construction season, then abandoned when the completed road provided access from below.
Their only remaining trace is in construction photographs, appearing as thin scratches on the slope above the road base, merging with the talus a few hundred feet above the camera frame. The road itself is the record. There is a section between the Weeping Wall and Logan Pass where the road base was cut into the Garden Wall at its steepest angles. The outer edge of the pavement ends at a natural stone lip reinforced with blocks of argillite, placed by construction workers in the late 1920s.
On the other side of that lip, the wall drops to the valley. If you stop at the parking lot just below Logan Pass, walk to the edge and look down, you will see the road line below you at the sharp switchback, and below that the Garden Wall slope, and below that the valley floor nearly two thousand feet down. The scale of the scene does not register immediately; it takes a moment to absorb what you are seeing. What you are seeing is a shelf in a cliff face, a shelf twelve feet wide and twelve miles long, carved by hand into a mountain that took 1.
7 billion years to become what it was before the men with drills arrived. They borrowed their workspace from the rock, took what it allowed, and shaped it with iron and powder, and with their understanding of how stone breaks and where it holds. Then they walked the access trail for the last time in October, and winter came and covered what they had built. And when the snowplow came up in June, the road was still there.
This scene repeats every June since 1933. Drill marks from the Garden Wall construction are still visible on some parts of the cut surface. Regular circles carved into the argillite where star drills entered ninety years ago. There are no signs or markers.
Most visitors pass them at fifteen miles per hour, enjoying the view without noticing them, but they are there. And if you slow down and look at the cut surface of the rock as you pass through the steepest part of the Garden Wall, you will see them. Small circles in the stone, precisely placed, the diameter of a steel drill bit driven by a man tied to a rope above a valley he was told was a national park and would be cared for forever. He drilled the hole.
He placed the explosives. He moved away, and the mountain fell, and the road was cut, and his name is recorded nowhere that can be found, but the hole is in the rock and the road is still there.
