In the first week of January 1938, an ironworker standing on the still-unfinished southern tower of the Thousand Islands Bridge looked eastward across the Saint Lawrence River and saw a mass of ice the size of a city block moving toward the construction barge below. It was coming fast enough that there was no time to move the barge or do anything else. The ice struck. The barge shook violently.

Men on deck clung to whatever was nearest—a cable, a mooring spike, a standing brace—and waited. The ice ground against the hull and kept moving, pushed by current and wind, until it disappeared around the bend of Wellesley Island. The men went back to work. The river did not stop.

Neither did the schedule. The ice was part of the job, the way the cold was part of it, the way the winds off Lake Ontario were part of it, and the way the Saint Lawrence in January was never a place designed for human construction work. Those conditions were the job itself. This is the story of how the United States and Canada built a bridge across the Saint Lawrence at the Thousand Islands in eighteen months, through two winters, against ice and current, and against the political complexity of a project that required the approval of two governments and four jurisdictions over something they had disagreed about since the 1890s.

It is the story of an engineer who understood what wind does to bridges before wind destroyed a bridge in a way the engineering world could not ignore. It is the story of a workforce that came to the Saint Lawrence from its last job site, worked in the river under conditions that the labor laws of the time barely regulated in any meaningful way, and left for its next destination the moment the work here was done. And it is the story of a dedication ceremony attended by the heads of two governments—but not by the men who built the bridge, because they had already gone. To understand why crossing at the Thousand Islands took forty years of proposals and rejections before a single foundation pier was driven, you need to understand what the Saint Lawrence River is and what the islands are, and why crossing that particular spot was harder than crossing a river between two provinces or two states.

The Saint Lawrence is the main outlet of the Great Lakes. It gathers the waters of Lake Superior, Lake Michigan, Lake Huron, Lake Erie, and Lake Ontario and carries them northeast through Quebec into the Gulf of Saint Lawrence and then into the North Atlantic. The volume of water it moves is enormous. At Kingston, Ontario, where the river leaves Lake Ontario and begins its run to the sea, the average flow is about 246,000 cubic feet per second.

The current that produces is not the gentle flow of a river through a flat farming valley. It is a continual force running through a channel shaped by ancient granite, at speeds that ships must calculate rather than ignore. The Thousand Islands begin where the river leaves Lake Ontario. The name is almost accurate.

The group includes about 1,800 islands, by a count that includes every rock that rises above water for at least one year out of three. Some are large enough to hold farms, summer estates, and private roads. Some are small enough to hold a single tree and a flagpole. The archipelago straddles the international border between the United States and Canada.

The boundary runs through the water itself, through the middle of the shipping channel, so that some islands are American territory and some are Canadian. The main navigable passage between them is an international line drawn by treaties through a geological formation that does not recognize political borders. The railroads reached the Thousand Islands in the 1870s from both sides. The Grand Trunk came up the Canadian shore.

The New York Central came up the American side. The summer resort industry, which had been growing in the region since wealthy families from New York and Toronto discovered the combination of clean water and dramatic island scenery in the 1870s, expanded quickly once the railroads cut travel time from the cities to hours. Hotels went up on the larger islands. Boat docks went in along the shores.

The steamboat excursion trade grew into a major seasonal business. All that movement—freight, tourists, residents, commerce between the two countries—crossed the Saint Lawrence at the Thousand Islands by ferry. There had been ferry service at this crossing since before the railroads, since before the resort industry, since before the automobile; the ferry was the only way across. By the second decade of the twentieth century, the car began changing the calculus of the ferry system, just as cars were changing the calculus of every ferry system in the country.

A ferry built to carry passengers and horses could be adapted to carry passengers and automobiles, but the volume of cars the road network was delivering to the ferry dock far outpaced the capacity of vessels designed for the pre-automobile era. New vessels were built. Capacity was expanded. Still the lines grew longer, because the number of cars was growing faster than any expansion of the ferry fleet could keep up with.

A driver arriving at the American ferry dock at Alexandria Bay on a Friday afternoon in the late 1920s might wait two hours before boarding. The ferry route from Alexandria Bay to Gananoque on the Canadian side was short. The problem was the dock, the vessel, and the schedule. These were not engineering problems.

They were capacity problems, and a bridge could solve them in a way no ferry expansion could. The solution everyone proposed was a bridge. A solution that had been proposed since at least 1890. The first serious formal proposal for a fixed crossing at the Thousand Islands went before legislators and transportation planners in the 1890s.

It came to nothing. Engineering challenges were cited. The river was deep in the shipping channel. The current was strong.

The winter ice buildup was heavy. The navigational clearance requirements for the large vessels of the Saint Lawrence trade meant that any fixed span across the main shipping channel would need very tall towers and a main span several hundred feet long—in the 1890s, a structure at the edge of what suspension bridge design technology could reliably provide. The political challenges were more complicated than the engineering ones and less easily solved by mathematics. A bridge across the Saint Lawrence at the Thousand Islands was not a bridge between two American states.

It was a bridge between two sovereign nations. Every decision about its design, construction, financing, and operation required agreement between Washington and Ottawa. The bridge authority required concurrent legislation in both countries. The financing structure required coordination between American and Canadian bond markets operating under different regulatory systems.

The customs and immigration facilities on each side required negotiations between federal agencies in two countries whose procedures were similar enough to invite comparison but different enough to require separate arrangements. The precedents for binational infrastructure were limited. The mechanisms for making the project work had to be built alongside the bridge itself. In the years between the first serious proposals and the mid-1930s, the crossing was studied, proposed, budgeted, argued over, and rejected.

Engineers came to the Thousand Islands and reported that the river was fast, the ice was heavy, the foundation conditions demanded attention, and the cost would be great. Politicians thanked the engineers and postponed action. The ferry lines grew longer. Complaints from resort owners and business interests on both shores grew louder.

The complaints were recorded. The postponement continued. What changed in 1934 was not the engineering. The engineering had been sufficient to build the bridge since at least the 1920s.

What changed was political agreement on both sides of the border, and more specifically, the creation of a financing mechanism that made the project possible without requiring either national government to commit public money to its construction. The Thousand Islands Bridge Authority was established by concurrent legislation in the United States Congress and the Canadian Parliament in 1934. It was a binational authority with members appointed by both governments. Its mandate was to plan, finance, build, and operate a bridge in the Thousand Islands.

It had the power to issue revenue bonds backed by future toll collections. The authority relied on neither national treasury. The bondholders bore the revenue risk. If the bridge earned enough in tolls to service the debt, the bonds were paid.

If it did not, no government would compensate the bondholders. This structure was politically acceptable in both countries because it insulated elected governments from the financial risk of a project that was controversial enough without the added possibility of a loss funded by public money. To sell bonds at prices the project could repay from toll revenue, the authority needed an engineering plan credible enough to convince bond buyers that the bridge would be built on schedule, opened, and generate the traffic projections required to meet the debt obligations. A credible engineering plan required a credible engineer.

The Thousand Islands Bridge Authority hired David B. Steinman. In 1936, when Steinman took on the Thousand Islands assignment, he was forty-six years old. He had grown up poor in Manhattan, in the shadow of the Brooklyn Bridge, the son of Eastern European immigrants who came to the United States with the same mixture of ambition and limited capital that brought most immigrants of that generation to American cities.

He watched the East River traffic from the Manhattan shore as a boy and decided that bridges were what he wanted to build. He took his engineering degree from Columbia University and a doctorate from the same institution, and spent the next two decades building bridges and consulting on bridge projects across North America and internationally. His record by 1936 included the St. Johns Bridge in Portland, Oregon; the Carquinez Strait Bridge in California; the Waldo-Hancock Bridge in Maine; and bridge commissions in South America and Europe.

He was not an unknown practitioner looking for a big project to prove his reputation. He was an established engineer with a track record of structures that performed as calculated. What distinguished Steinman from other engineers who could have taken the Thousand Islands assignment was not his list of completed bridges. It was what he had been writing and arguing about for the previous decade concerning bridges that had not yet collapsed.

Steinman had been publishing technical papers since the late 1920s arguing that the trend in suspension bridge design toward thinner, lighter, more economical decks was creating bridges increasingly vulnerable to a mode of failure the engineering community was not adequately accounting for. He called it aerodynamic instability. Other engineers dismissed his concerns as theoretical. The bridges built to the light-deck standard were standing.

They were carrying traffic. The arguments against thin decks were about what might happen, not what had happened, and arguments about what might happen are less persuasive than bridges that exist. Steinman designed the Thousand Islands bridges with stiffened trusses. The trusses were designed to carry real mass and real rigidity.

This was not fashionable by the standards of suspension bridge economics in 1936. It was sound. He delivered the engineering plan to the authority in 1936. The bonds were sold.

Construction was announced for the spring of 1937. The geography of the Thousand Islands crossing did not allow a single bridge to be strung from shore to shore. The archipelago cut the route in such a way that a conventional suspension bridge with two towers and a single main span was not possible unless the span was so long that the cost and engineering complexity would make the project unviable. Steinman’s design used the islands as intermediate supports and divided the crossing into a series of bridges connected by roads on the islands.

The route crossed from Collins Landing on the American shore to Wellesley Island, the large American island near the middle of the archipelago. From Wellesley Island, the route crossed the international boundary channel—the deeper, faster water used by the large Saint Lawrence vessels—to Hill Island on the Canadian side, where the international boundary line intersects the highway. From Hill Island, the route continued to the Canadian mainland at Lansdowne, Ontario. The two main suspension bridges were the American crossing between Collins Landing and Wellesley Island, and the Canadian crossing between Hill Island and the Canadian mainland.

Each main bridge required its own pair of towers, its own cable system, and its own anchorages. Between and beside these two main spans, approach bridges of simpler design connected the road sections on the islands to the suspension bridges and to the shores. The total highway length across the crossing, including the road sections on the islands, was about eight and a half miles. The bridge structures themselves were shorter.

Each main span was about 800 feet between towers. The towers rose about 235 feet above the water. Compared to the major suspension spans of the era—the George Washington Bridge with its 3,500-foot main span, the Golden Gate at 4,200 feet, the Mackinac Bridge that Steinman would later design at 3,800 feet—the Thousand Islands bridges were modest. By the standards of what the crossing’s geography required and what the navigational constraints of the Saint Lawrence passages permitted, they were exactly what the site demanded.

Construction began in the spring of 1937. The men who worked on the Thousand Islands Bridge in 1937 came from the same labor pools that had been supplying American bridge construction projects since the end of the previous century. The ironworkers came from Kahnawake. The Mohawk community on the south bank of the Saint Lawrence opposite Montreal had been sending men to build steel in high places since 1886, when Kahnawake ironworkers were first hired for the dangerous work on a railway bridge across the Saint Lawrence at Sault Ste.

Marie. The story the ironworkers tell, and the records support, is that the Kahnawake men walked the bare steel without hesitation when the first iron bridge was stripped of its temporary scaffolding, and the construction supervisor noticed and hired them specifically for the high work others found difficult. In the decades that followed, ironworkers from Kahnawake went to build the steel rising in New York, Chicago, Pittsburgh, and Montreal. They worked on the Empire State Building.

They worked on the Washington Bridge. Their reputation was built on something real—a specific combination of physical balance, spatial awareness, and comfort with exposure to heights that allowed a worker to function at full mental capacity when the steel was moving, the wind was blowing, and the ground was very far below. The Thousand Islands was, in a sense, homeland for the Kahnawake men in a way the Straits of Mackinac or the Hudson River crossing were not. The Saint Lawrence was their river.

Their community sat on it. These men had grown up watching the river and knew its ice, its currents, and its seasonal behavior the way people who live on a river know it—not from engineering reports, but from direct experience through every year of their lives. They came to the Thousand Islands job site with that knowledge in addition to their steel-working skills. They lived in boarding houses in Alexandria Bay and Gananoque during construction.

Some brought their families; most did not. The work was scheduled to last into the following summer, and the community they came from was across the river and up the shore, but it was not where they slept for the duration of this job. They worked as ironworkers work on any project. The work was the objective, and home was waiting when it was done.

The marine construction workers came from among the men who follow underwater foundation work from project to project. The Saint Lawrence imposed its own conditions on them. The cofferdams had to be placed and seated in a current moving at several knots in the main shipping channel. Lowering a cofferdam to the riverbed in a strong current required mooring systems and careful calculations that work in still water never demanded.

The barges holding the cofferdams in position required anchor chains with greater load capacity than still water, with extra weight needed to resist the constant push of the current. The geology at the tower foundation sites added another layer of difficulty. The Saint Lawrence at the Thousand Islands runs through the Frontenac Arch, a granite formation among the oldest exposed bedrock in North America. The Laurentian Shield granite that forms this formation is not soft material.

Drilling foundations in granite is fundamentally different from drilling in the lakebed sediments that the Mackinac foundations would encounter seventeen years later. The cutting edges on the cofferdams required equipment capable of removing hard rock rather than displacing soft sediment. The digging was slower. The work in the compressed-air chambers at the cutting edge of the cofferdams was done against material that did not move easily, requiring continuous mechanical effort.

Working depths in the main shipping channel ranged from 50 to 70 feet below the water’s surface. At these depths, the air pressure required to keep water out was far less than the pressure required for the 200-foot-deep Mackinac foundations, but it was not easy. A man working a full shift at sixty feet in compressed air absorbs nitrogen into his blood and tissues, requiring staged decompression on the way up. The United States Navy decompression tables of the time set the stop periods and the time required for each based on depth and duration of work.

The construction companies used those tables. The medical staff on site knew the signs and what to do when decompression sickness symptoms appeared. The protocol relied on the men following it, and the men following it relied on their understanding of the consequences of not doing so. In a work environment where the schedule was the primary pressure, those consequences were not always explained as thoroughly as the medical staff would have preferred.

A man who had worked a full shift in compressed air and cold water below the surface, wanting to get up and eat something warm, was liable to have his judgment about how fast to ascend affected by his exhausted physical state. The injuries that resulted from compressed-air work on the Thousand Islands foundations are recorded in the construction companies’ records. They are not in any document widely available to the public. The work was completed to specification.

The caissons sit in the granite of the Frontenac Arch, poured and sealed, carrying the loads the bridge has imposed on them since 1938. They have never moved. The construction crews worked on the surface of the Saint Lawrence through the winter of 1937–1938 in conditions that no modern safety regulation would permit at an outdoor construction site. The average temperature at the Thousand Islands during January and February was in the single digits Fahrenheit.

It dropped to twenty below zero or worse when cold air masses came down from the Canadian interior without the moderating effect of lake moisture to keep the temperature from falling without limit. The wind that accompanied the low temperatures came off Lake Ontario from the west through the Thousand Islands channel unobstructed, meaning there was nothing between the water and the construction platform to slow it. Wind at zero degrees Fahrenheit is not wind at thirty degrees. The effective temperature felt by exposed skin in those conditions is lower than the measured temperature by enough that skin loses function in minutes and freezes in the time it takes to become seriously concerned.

The men wore wool coats, leather gloves, and canvas jackets over insulating clothing. The gloves that kept their hands from freezing made the precise work of bolting, welding, and rigging slower than bare hands would have allowed. That was the choice: functional but imprecise hands, or bare hands that worked precisely for a short time and then stopped working entirely. The men wore the gloves.

Steel at twenty below zero is steel whose thermal contraction has tightened every bolt and closed every tolerance. A bolt that turns smoothly at fifty degrees requires more torque at twenty below. A connection that fits precisely in warm weather requires adjustment in extreme cold because the steel has changed dimensions. The ironworkers making connections on the rising towers during the Thousand Islands winter were handling a material that behaved differently than it was designed for, wearing gloves, at heights where the wind was faster and colder than at the water’s surface, where the consequences of a mistake were always the same when there was no net beneath you.

The Saint Lawrence does not freeze solid in a Thousand Islands winter. The current through the main shipping channel prevents stable ice from forming across the channel’s full width in most years. What forms instead is an ice regime in various states. Anchor ice grows on the riverbed and breaks loose when the current undercuts it.

Frazil ice forms in the water column when supercooled water crystallizes around suspended particles, creating a moving slurry that builds up against obstacles, barges, and piers. Shore ice extends from the banks in cold snaps and breaks off in large sheets when the temperature rises slightly and the current finds a way under the ice from below. The ice floes moving through the Thousand Islands from Lake Ontario posed a danger to the construction barges—intermittent, unpredictable, and sometimes large enough to force the entire crew to stop what they were doing and hold on to something. The moorings of the construction barges were sized for the expected current load plus a margin.

Ice floes were not in the mooring calculations, because the size of any given floe on any given day could not be predicted in advance. When the floe exceeded the mooring margin, the moorings dragged. When the floe was big enough, the anchor chain parted. When the chain parted in the middle of a crane lift, the material being lifted had to be brought back down to the barge deck before the barge could be repositioned, re-moored, and the lift resumed.

These interruptions were not in the construction schedule. The schedule was set by the work defined in the engineering drawings, not by the conditions the Saint Lawrence imposes in January. The schedule did not change. The crews worked faster when the ice was not running, stopped when it was, and went back to work once it had passed.

The barges had heated shelters below deck where men could go during breaks to bring their body temperature back to a level that allowed them to continue working. These shelters were not heated in the usual sense of a warm room. They were spaces warmer than the deck, the way any enclosed space is warmer than the open air in a northern river winter. They were enough to prevent hypothermia over the course of a break.

They were not comfortable. Comfort was not a standard the conditions allowed. The men who fell into the Saint Lawrence in winter did not stay in it long enough for cold shock to be the cause of death. The water temperature in January was near zero Celsius.

Immersion in water that cold triggers an involuntary gasp reflex that overrides voluntary breath control. A person who falls into the Saint Lawrence in January with his head underwater when the gasp occurs inhales water. Drowning before hypothermia set in was the outcome in construction accidents in the river that winter. The construction crews kept safety lines and rescue boats with equipment to get a man out of the water as quickly as the crew could respond.

The response time was as fast as human reflexes and the river current allowed. In the current, in the ice, and in winter, that was not always fast enough. The deaths on the Thousand Islands Bridge construction are in the accident records of the construction companies and in the files submitted to the relevant regulatory bodies in New York State and Ontario. They are in the records and not in circulation among the public.

The families of the dead received settlements negotiated through the workers’ compensation frameworks of that era in both countries. Frameworks that provided limited compensation by the standards that would develop in the following decades. The construction companies and their insurers presented these settlements as adequate. Whether they were adequate is a question the records were not written to answer.

By the spring of 1938, the foundations were in place and the pier columns had risen above the river’s surface. Tower construction had continued through the winter on both main bridges, and by the time ice conditions in the shipping channel allowed full resumption of barge operations in April, the towers on the American and Canadian sides were approaching their full height of about 235 feet. The towers rose in sections lifted by cranes mounted on barges. Each tower section was a fabricated steel structure made at a plant on the mainland, moved to the site by barge, transferred to the barge crane at the work site, and lifted into position onto the rising tower column.

The ironworkers at the top of the rising tower received the section, guided it into place using guide lines and their hands and bodies, and made the bolted connections that secured the new section to the one below while welding crews followed to seal the connections. At 235 feet, the completed towers were the tallest structures in a region where no permanent structure exceeding three or four stories had ever been built. From the cable saddles on a clear spring morning, a man could see both banks of the Saint Lawrence, the American shore to the south and the Canadian shore to the north, the Thousand Islands archipelago stretching in both directions along the river, and Lake Ontario to the west, where the river begins and where the water comes from, with the river flowing east toward Montreal and the sea. The ironworkers who climbed to the tops of the towers to set the cable saddles did not stay up there longer than the work required.

There was more work below, and the schedule did not change just because the view was spectacular. Cable spinning on the Thousand Islands main bridges began in the spring of 1938 as the tower work neared completion. The process was the same one John Roebling had developed in the previous century, refined on every major suspension bridge built since. A spinning wheel carrying wire from a reel at one anchorage was driven across a footwalk by a worker walking from bank to bank over the tower tops.

At the far anchorage, the wire was secured. Then the wheel returned empty. The process began again with the next wire. This continued until the full complement of wires was in place, then the bundle was compacted and wrapped.

The footwalks on the Thousand Islands bridges were not five miles long like the Mackinac Bridge’s. The main spans were about 800 feet. The footwalks were correspondingly shorter, but they remained wire mesh strung bank to bank at heights up to 235 feet above the river, and the mesh flexed underfoot, and the river ran below, and the wind coming off Lake Ontario crossed the channel between the islands in a way that moved the mesh sideways under a man’s feet. A worker going up the walkway for the first time in the spring of 1938 had to find his balance on a surface that was not solid ground and moved in ways ground does not move.

A worker who had spent three weeks on the walkway walked it the way he would walk a dock, without looking down unless the work required it. The cable wire for the Thousand Islands was high-strength galvanized steel. The galvanizing process protected the steel from the moisture and the freeze-thaw cycles that the Saint Lawrence climate would impose on the cables over decades of bridge service. Spinning continued in shifts through the spring and early summer of 1938, with the wheel making its journeys between the anchorages as many times as possible on each working day as the crew could drive it.

By summer, the cable bundles were completed, compacted, and wrapped, strung from the anchorages over the cable saddles at the tower tops and in drooping curves between the towers down to the anchorages on the distant shores. They were ready to carry the weight of everything that would hang from them. The deck sections came from the fabrication plants and arrived at the site by barge. The floating cranes lifted each section off the barge deck and swung it into position between the suspender ropes hanging from the main cables.

The sections were connected to the suspenders and to each other, and the deck extended from each tower toward the midpoint between the towers. When the two halves coming from the opposite towers met in the middle, they were connected. The alignment was within the tolerances Steinman had calculated. The mathematics had been precise.

The work had followed the mathematics. The approach bridges connecting the main spans to the road sections on the islands and the shores were completed during the summer. The road surfaces were paved. The toll plaza facilities and the customs and immigration inspection stations were built and equipped at both the American and Canadian ends, because a bridge between two nations requires the border-control mechanisms that a bridge between two states does not.

On August 18, 1938, the Thousand Islands Bridge opened to traffic. Franklin Delano Roosevelt, President of the United States, was there. William Lyon Mackenzie King, Prime Minister of Canada, was there as well. Two heads of government at the dedication of a single bridge—not a common thing, reflecting the binational character of what had been built and the political importance both governments decided to attach to it.

Roosevelt spoke of the bridge as a symbol of friendship between two neighboring nations. King spoke of the practical meaning of the crossing for trade and for people on both sides of the river. David Steinman was there. His name is in the ceremony program.

His name is in the newspaper coverage that ran in the papers of both countries. The ironworkers from Kahnawake who built the towers and deck sections were not in the crowds at Collins Landing on August 18, 1938. The marine construction workers who set the caissons into the face of the Saint Lawrence current were not in the crowds. The laborers who moved materials and ran the equipment and did the work that never appears in engineering specifications were not in the crowds.

They were where construction workers go when a project ends. They were at the next project. The next project was somewhere else, and the work on the Saint Lawrence was finished, and finished work does not require its men to stay beside it. The ribbon was cut.

The cars crossed. The bridge was open. Two years and three months after the Thousand Islands Bridge opened to traffic, on November 7, 1940, the Tacoma Narrows Bridge in Washington State twisted until it broke apart in a wind of 42 miles per hour and fell into Puget Sound. The failure was the most important event in suspension bridge engineering in a generation.

The Tacoma Narrows deck had been designed to be thin, light, and economical—the characteristics that made it vulnerable to the aerodynamic forces a bridge over Puget Sound could encounter. The bridge oscillated. The oscillations grew. The structure exceeded its capacity to absorb the oscillating loads and came apart.

The Thousand Islands Bridge, opened two years earlier and designed by the same engineer who had spent a decade arguing that thin, lightweight decks create aerodynamically unstable structures, stood without problems on the Saint Lawrence. It did not oscillate. The stiffened truss that Steinman had insisted upon—heavy by the standards of 1930s economical design philosophy—worked exactly as he had predicted. The Tacoma Narrows failure proved to the engineers who had not been convinced by Steinman’s arguments before November 7, 1940, that those arguments were correct.

The Thousand Islands Bridge was one of the structures built on the correctness of those arguments before the evidence arrived. Steinman went on to design the Mackinac Bridge, which opened in 1957. The link between the two projects is direct. The engineer who developed his aerodynamic design principles on the Saint Lawrence in 1938 applied those principles at the Straits of Mackinac in 1957, and the Mackinac Bridge has stood for decades in Great Lakes weather without oscillating, as the Thousand Islands Bridge has stood for decades in Saint Lawrence weather without oscillating.

These are not coincidences. They are the same solution to the same problem, built by the same engineer, proven first on a smaller river and then on a larger strait. The Thousand Islands Bridge Authority was not dissolved after the opening. It continued to operate the crossing and use the toll revenue for the maintenance and capital improvements of the structure.

The revenue bonds that financed construction were repaid from toll income over the years following the opening. The bridge paid for itself from the tolls of the people who crossed it, which is what the bond structure required, and traffic exceeded the projections presented to the bond buyers when they purchased. Maintaining a bridge on the Saint Lawrence is a permanent operation. The river’s freeze-thaw cycles, ranging between summer and winter temperatures every year, impose stress on the concrete, steel, paint, and insulation, requiring continuous inspection and repair.

The cables are inspected on a set schedule. The deck is examined for stress in the welds and bolted connections. The tower columns are inspected for corrosion in the splash zone where the river surface meets the steel. The suspender ropes are checked for broken wires.

The road surface is maintained. The work never stops. A bridge that does not receive continuous maintenance is a bridge aging toward a state where the cost of restoration exceeds the cost of replacement. At a vital crossing like the Thousand Islands bridges, part of the road network connecting the northeastern United States to eastern Canada, that situation is not acceptable.

The bridge is 87 years old as of this writing. The tower bases set by the ironworkers from Kahnawake in 1938 are still in place. The caissons that the compressed-air workers set into the granite of the Frontenac Arch are still seated in the granite. The cables spun by workmen walking the wire footwalks in the spring winds of 1938 are still in their wrappings, inspected and monitored, unchanged in their essential structure since the day the last wire was placed.

If you cross the Thousand Islands Bridge today, you cross the international boundary between the United States and Canada on the Canadian main bridge, at the point above the main shipping channel of the Saint Lawrence where the current runs at several knots toward the sea. The Frontenac Arch granite lies under the riverbed, under the caissons, under the concrete seals poured by the compressed-air workers in 1937 who dug through the hardest bedrock in North America to reach the bearing material the engineers specified. The tower above you is 235 feet of steel assembled in sections over two winters, bolted and welded by ironworkers in leather gloves at temperatures cold enough to shrink the steel from its design dimensions and tighten every bolt in place. The cable above the tower was spun wire by wire across a mesh walkway over the river in the spring winds of 1938 by men who learned to walk that surface without looking down, because looking down while working was not compatible with the task.

The bridge deck under your tires was fabricated on the mainland, shipped by barge to the site, and lifted into place by cranes on barges that the Saint Lawrence current pressed against continuously from the moment the moorings were set until the moment the last deck section was connected and the barges left. All of it was done in eighteen months. Two winters on the Saint Lawrence, and the bridge was ready. Roosevelt’s name is in the ceremony record.

King’s name is in the ceremony record. Steinman’s name is on the engineering drawings, in the commemorative literature, and in the history of American suspension bridge design. The ironworkers from Kahnawake who assembled the towers in winter temperatures on a river they had grown up beside—their names are in the payroll records. The compressed-air workers who set the caissons into the granite—their names are in the accident records.

The laborers who moved materials and ran the equipment and did what construction requires that never appears in engineering specifications—their names are in the payroll records. None of them are on the bridge. The bridge they built is on the road maps and in the tourist materials of two countries, and on the windshields of every driver who has crossed it in 87 years of operation. A crossing that took forty years of proposals to begin took eighteen months to complete once the money was available, the schedule was set, and the men arrived at the river.

The Saint Lawrence still runs. The current still moves at several knots through the shipping channel. The ice still forms along the channel edges in January and moves across it when the current undercuts it. The granite of the Frontenac Arch has not moved.

The bridge is still there.