In the autumn of 1890, sixty feet beneath the bed of the St. Clair River, on the American side of the international boundary, a compressed-air worker—listed on the payroll of the Michigan Central Railroad as a “sand hog”—held his post at the face of a tunnel that would not be finished for another six months. He directed a pneumatic drill into clay that nothing had touched since the glaciers had laid it down twelve thousand years before. The air around him was compressed to twenty-six pounds per square inch above the pressure of the atmosphere at the surface.

The river above ran at its autumn level, about six feet over the crown of the tunnel. The clay around him held. The compressed air was what kept it from collapsing in. The men at his sides were digging at the same pace, cutting through the same clay, breathing the same compressed air, and earning the same differential over the ordinary labor rate that the railroad paid to men who worked under pressure, because men who worked under pressure did not always come back to the surface the same way they had gone down.

He did not know, the way a man working underground in 1890 could know, whether the clay overhead would hold through his entire shift. He knew what the engineers had told the supervising foreman, and what the foreman had passed on: that soil samples taken along the tunnel’s line had shown blue clay of consistent character at the working depth—clay that compressed air could hold against the water of the river above. He knew what the men who had worked before him knew: that the clay was mostly consistent, and “mostly” was the word that kept a man careful. He worked his drill and moved the cut material into muck cars on a light railway that ran back toward the shaft to be hauled to the surface.

He breathed the compressed air, and he did not come up too fast at the end of his shift, because men who came up too fast got the bends, and men who got the bends badly did not finish the job. The St. Clair Tunnel opened on September 19, 1891. It was the first underwater tunnel in North America built by the compressed-air shield method.

It ran 6,025 feet from Port Huron, Michigan, on the American side to Sarnia, Ontario, on the Canadian side. Its bore was nineteen feet ten inches in diameter, large enough to pass standard-gauge railroad cars of that era without altering their dimensions. It connected the Grand Trunk Railway’s main line, running east from Toronto through southwestern Ontario, with the Michigan Central system, which ran west from Detroit through Port Huron to Chicago and the western rail network. The link eliminated a ferry that carried railroad cars one at a time across a river that froze in winter, ran ice in spring, and produced fog in autumn, making the schedule of every freight shipper who crossed that border uncertain in a way a fixed link under the river would not be.

The men who built the tunnel exist in the payroll records of the Michigan Central Railroad and the contractors it hired. The payroll records exist in archives, where the records of that era remain in whatever condition a century of institutional custody left them in. Their names are not on the tunnel entrance at Sarnia. They are not on the entrance at Port Huron.

They are not on anything that would catch the eye of a person crossing the border on the Blue Water Bridge today, looking down at the river from the highway span. This is the story of how they drove that tunnel under a river between two countries, what it cost the men who drove it, and why both halves of that question are worth knowing. To understand why crossing the St. Clair River at Port Huron was a specific problem, and why it waited so long for a permanent solution, you have to understand what the St.

Clair River is, and what it does in the continental geography it occupies. The St. Clair is not a river the way the Mississippi is a river. It does not drain a watershed in the traditional sense.

It is an outlet. It is the channel through which Lake Huron empties into Lake St. Clair, which empties through the Detroit River into Lake Erie, then into the Niagara River, Lake Ontario, and finally through the St. Lawrence to the Atlantic.

The St. Clair River is about forty miles long. It runs south from the southern end of Lake Huron at a width between 1,600 and 4,000 feet. Its current is not the current of a rain-fed, snowmelt-fed river.

It is the current of Lake Huron draining—steady, continuous, running at four to five miles per hour through the navigation channel at the strait at Port Huron, where the lake narrows into the river and the full volume of the lake’s discharge accelerates through the reduced cross-section of the channel. At Port Huron and Sarnia, the river is about 2,000 feet wide across the navigation channel. The depth of the main channel there runs between twenty and twenty-five feet. The bottom is the blue clay the glaciers left when they retreated.

Dense and moldable, at the depth of the tunnel crown it was coherent enough that the engineers who designed the crossing believed it could be held against water seepage by compressed air at pressures the human body could tolerate for the length of a shift. In the 1850s, the Canadian Grand Trunk Railway reached the Ontario side of the St. Clair at Sarnia and had to cross to the Michigan shore. The answer was the ferry.

Car ferries on the St. Clair ran into the same problems that car ferries faced at every major river crossing: weather, ice, fog, capacity, and schedule. The St. Clair freezes.

Not every winter, but in a cold winter the freeze could stop the ferry for weeks. Ice running in the river during the spring thaw could disable a vessel or damage it. Fog forming over the river when warm water met cold air could close the crossing for a full day at a time. A shipper whose goods waited on the Michigan side of the river for the fog to lift paid for time the railroad had promised he would not have to pay for.

The Grand Trunk and the Michigan Central, which connected with the Grand Trunk at the St. Clair crossing, studied building a bridge. The current was manageable for bridge pier construction. The depth was not extreme.

The width was not an obstacle to the bridge technology of the 1870s and 1880s. What was the obstacle was navigation. The St. Clair carries the upper Great Lakes trade to the lower lakes and the St.

Lawrence—bulk carriers moving iron ore from the Lake Superior ranges to the steel mills of the Ohio Valley, grain ships moving wheat from the prairie ports of Lake Michigan and Lake Superior to eastern mills, package freight vessels running on schedules between lake ports. A bridge across the St. Clair at Port Huron required a navigation clearance dictated by that traffic, which the bridge technology of the age could not provide except at a cost the railroad economics of a single-line crossing would not justify. The bridge would have demanded a main span high enough to pass loaded ore carriers and grain ships—which were growing larger through the 1870s and 1880s as the lake trade developed—plus a movable span or lift mechanism for vessels needing passage at heights beyond the fixed structure.

And the navigation disruption caused by any movable part of a bridge was exactly what the river’s traffic was trying to avoid. The tunnel was the solution the engineers arrived at, independently and together, in the late 1880s. A tunnel under the river solved the navigation problem by going beneath it. It did not obstruct marine traffic.

It required no drawbridge. It was not subject to ice and fog the way the ferries were. It was, as the engineers put it before the railroad’s board in 1886, a permanent solution to a problem the ferry had been managing temporarily for thirty years. The engineering firm charged with designing the crossing was that of Joseph Hobson, chief engineer of the Grand Trunk Railway, who had built bridges across the river systems of Ontario and had studied the tunneling methods being developed in Europe and applied with mixed results in American railroad tunnel projects of the 1870s and 1880s.

Hobson’s firm studied the St. Clair crossing and concluded that the shield tunneling method developed by James Henry Greathead in London, modified by American engineers for particular soil conditions, was applicable to the blue clay of the St. Clair riverbed at the depth the tunnel’s line required. The Greathead shield was a cast-iron cylinder pushed forward by hydraulic jacks braced against the completed tunnel lining behind it, with a cutting edge at its front face that bit into the material ahead.

Men working at the face inside the shield excavated the material ahead of the cutting edge. The shield advanced as material was removed. Behind the shield, cast-iron rings were bolted together to form the permanent lining of the tunnel. The process was continuous.

The face was always advancing, always being excavated, always being lined behind the point of advance. What made the St. Clair crossing possible—what made the Greathead shield work in London clay and in the American tunnel projects that followed—was compressed air. At the depth of the St.

Clair tunnel crown, about fifty feet below the river’s surface and twenty feet below the riverbed, the hydrostatic pressure of the water above was about twenty-two pounds per square inch. Pumping compressed air into the working space at the shield face and throughout the tunnel under construction, at a pressure slightly above the hydrostatic pressure of the river, pushed the water back. The clay tunnel face stayed dry. Men could work at it.

The compressed-air plant for the St. Clair was assembled on both banks of the river. Steam-engine-driven compressors pushed air through pipes into the tunnel below, through air locks that controlled entry and exit from the pressurized zone. The air lock was a steel cylinder with two airtight doors—one opening to the surface atmosphere, one opening to the pressurized tunnel.

A worker entered through the surface door into the lock. The door closed behind him, and the air pressure in the lock was raised gradually to equal the tunnel pressure before the inner door was opened. A worker leaving the tunnel reversed the process, entering the lock from the pressurized side and spending the required decompression time as the pressure was lowered to atmospheric in controlled stages. The decompression protocol was the engineering specification that separated men working in compressed air from decompression sickness.

Engineers in 1890 understood the physics generally, even if the biochemistry was not fully characterized until decades later. The nitrogen in compressed air dissolves into the blood and body tissues under pressure at a rate proportional to the pressure and the duration of exposure. If the pressure is lowered slowly enough, the dissolved nitrogen returns gradually to the gaseous state within the lungs and is exhaled without forming bubbles in the tissues. If the pressure is lowered faster than the lungs can clear it, it comes out of solution in the blood and in the tissues—in the joints, in the muscles, in the spinal cord, in the brain.

The pain it caused in the joints gave the condition its common name among the men who suffered it. Cases that reached the spinal cord caused paralysis. Cases that reached the brain were fatal. The specific decompression schedules for the St.

Clair Tunnel in 1890 were based on the best practice available at the time, derived from the work of the Brooklyn Bridge caisson workers in the 1870s, the Eads Bridge, and the smaller tunnel projects that followed. The schedules required men working at St. Clair tunnel pressures to spend forty to sixty minutes in staged decompression per shift, which limited each man’s actual working time and meant the number of men required to maintain the desired rate of advance was greater than a simple calculation of face area and excavation rate would suggest. The men who entered that air lock at Port Huron in the winter of 1890–91 faced the conditions of the Great Lakes region in those months—conditions imposed by Lake Huron, which stretches across 23,000 square miles of open water that does not freeze in most winters, on the Michigan shore.

Port Huron’s average temperature in January was in the low twenties Fahrenheit. The winds off the lake carried a particular cold, the cold of air that had crossed open water for two hundred miles without obstruction. A man walking to the head of the shaft at dawn in January 1891 in Port Huron walked through a cold that the brick building of the shaft-head machinery did little to block, and that the compressed air below did nothing to relieve in terms of warmth, because compressed air at twenty-six pounds per square inch above atmospheric pressure is not warm air. It is dense air, and dense air in a tunnel during a Michigan January is cold air under high pressure, with specific effects on a body already working hard against clay at the face of a huge shield in the dark.

The pay differential the Michigan Central paid to compressed-air workers on the St. Clair project was a premium above the ordinary labor rate of the era. The ordinary labor rate in the Port Huron area in 1890 was roughly $1. 25 to $1.

50 a day. Compressed-air workers received a premium of twenty-five to fifty cents a day above that rate. The premium reflected the nature of the work. The men who did it understood the premium in the terms available to workers in 1890—terms of what the work paid against the alternatives, and what the risk was against the risk of having no income.

The alternatives for an ordinary worker in Port Huron in 1890 were farm labor, work on the harbor docks on the river, or the small manufacturing that the city’s economy provided. None of those alternatives offered the compressed-air premium. The premium was real money in the economy of a working family in 1890. The bends were real too.

The number of cases of decompression sickness on the St. Clair Tunnel project, and the number of those cases that resulted in permanent injury or death, are not numbers that appear in any of the widely read accounts of the project. The engineering literature of the era documents the project as a technical success—the first underwater tunnel in North America, completed using the shield and compressed-air method, finished on schedule, within the cost estimates the engineers had given the railroad’s board. The engineering literature does not count cases of the bends because that count was not an objective of engineering literature.

The objective of engineering literature was to describe the method and its results for the benefit of engineers who would use it in later projects. The men who were injured exist in the medical records kept by the project physician on site, if such records exist, which is uncertain for a project of that era. They exist in whatever correspondence the project contractors filed reporting accidents to the railroad, if that correspondence survived in the company archive—which, for the Michigan Central of 1891, requires knowing where the company’s records ended up after the twentieth-century mergers and acquisitions that folded it into Penn Central and beyond. Tracing the man injured by decompression sickness at the face of the St.

Clair Tunnel in February 1891 to any surviving record is a path through an institutional history that most labor historians have not had the resources to fully follow. Joseph Hobson presented the completed tunnel to the Grand Trunk Railway in September 1891. The first locomotive passed through it on that occasion. Hobson received the recognition due an engineer of achievement—considerable recognition in the professional culture of 1891: papers read before engineering societies, entries in the technical encyclopedias of the era, a place in the profession’s record.

The contractors who managed construction on both the American and Canadian sides received their payments and their place in the construction records. The men who dug the tunnel exist in the payroll records, in the daily attendance books, in whatever residue of the project’s working papers survived the closing of the site. What they built changed the commercial geography of the Great Lakes region. Before the tunnel opened, the Grand Trunk’s main line from Montreal and Toronto to Chicago was divided at the St.

Clair River into two sections connected by a ferry subject to ice, fog, and capacity limits that traffic had already exceeded by 1885, when the tunnel idea was first seriously proposed. After the tunnel opened, the line was continuous. A through freight car could move from Montreal to Chicago without being broken down, loaded onto a ferry, and re-broken on the other side. The schedule promised to the shipper became the schedule the railroad could now keep, because the tunnel, unlike the ferry, did not stop for fog.

The tunnel served the Grand Trunk and its successor companies for sixty years. By the 1950s, the railroad cars carrying North American freight traffic had grown larger in every dimension—longer, wider, heavier than the cars the 1891 tunnel was designed to accommodate. The Greathead shield’s nineteen-foot-ten-inch diameter had suited the equipment of 1891, but it no longer suited the equipment of 1955. The Canadian National Railway, which had acquired the Grand Trunk in 1923, built a new tunnel beside the original in 1994 to accommodate modern equipment.

The 1891 tunnel went out of service. It remains in the riverbed, sealed. The 1994 tunnel is the one that carries railroad freight under the St. Clair today.

At Port Huron, the south portal of the 1891 tunnel can be seen from the riverbank—a brick arch sealed at the edge of the railroad right-of-way that borders the St. Clair. A historical marker stands there. The marker describes the tunnel, its dimensions, its date of completion, and the engineer Joseph Hobson.

It does not name the compressed-air workers. It does not describe decompression sickness—the bends. It does not describe the winter mornings at the shaft head, or the decompression protocol, or the pay differential, or what that differential meant to a family living in Port Huron in 1891 on the income of an ordinary laborer plus what the risk of working under pressure in January in Michigan added to the daily wage. It is worth being precise about what this omission is and is not.

It is not a failure of the marker in any individual sense. A historical marker is a small object in a public space and can carry only so much. What it carries is the general frame—the history, the engineer, the technical achievement. The general frame is accurate.

But the frame is also a process of selection, a selection based on what the available record gave priority to, and the record of the St. Clair Tunnel, like the record of every project of its era, gave priority to the engineer, the railroad, the date of completion, the technical first. The worker’s name existed in the daily attendance book, a document no one expected to survive. The St.

Clair still runs between Port Huron and Sarnia. The current still moves at four to five miles per hour through the navigation channel. The lake ships still pass there—bulk carriers moving iron ore, grain, and limestone, now far larger than the ships of 1891, clearing the bottom with margins allowed by the channel depth and monitored by the Coast Guard. The Blue Water Bridge, now two spans, carries highway traffic between Michigan and Ontario at the same crossing where the ferry once ran and where the tunnel passes beneath.

On a clear day from the bridge, you can see downstream where the river widens toward Lake St. Clair, and upstream where it narrows toward the southern end of Lake Huron. You cannot see the tunnel. It lies fifty feet below the riverbed, a hundred feet below the river’s surface.

The 1994 tunnel is still in service. The 1891 tunnel is closed and dark, still in the blue clay, in the place where the Greathead shield reached it foot by foot, while the men at the face worked their drills and moved the cut material into the muck cars, breathed the compressed air, and came up through the decompression lock at the end of their shifts with varying results. Some of their names are in the payroll records. Some of the payroll records are in the archives.

The archives are open to researchers with the knowledge to find them, the time to read them, and the interest to search them. The tunnel is there in the dark under the St. Clair, in the clay that held. It held because someone pushed a great shield through it foot by foot, came up through the air lock at the end of each shift, was paid, and left no trace in any record that a person looking at the Blue Water Bridge today would know how to find.

It is there in the clay. The clay held.