On a spring morning in 1935, a riveter standing on a wooden plank four hundred and thirty feet above the surface of San Francisco Bay looked east toward Oakland and saw fog. He looked west toward San Francisco and saw fog as well. The bay beneath him was not visible. The water was there, one hundred feet deep, a stretch of open bay with tidal currents running four knots through the main channel; but the morning fog, which had been crossing the Golden Gate since before any European ship entered the bay, had settled across its full width, and the plank the man stood on was immersed in it.
The top of the tower rose ninety feet above him into airc with no horizon. He was working between two invisible surfaces, performing a job that required him to perform at height under conditions that could not be described as favorable. He was part of a riveting gang. There were four men in the gang.
He was the riveter with the pneumatic hammer; the man behind the steel, called the bucker-up, who held the rivet in place from the other side of the surface using a holding tool; the catcher, who received the hot rivet from the heater and drove it immediately into the hole; and the heater, who worked at a portable coal-fired furnace twenty feet away, pulling the rivet from the fire when it reached the orange heat it had to reach before it would fill the hole it was driven into, then throwing it in an arc to the catcher, who caught it in a bucket or in his bare hand, depending on his preference and the thickness of his skin, and drove it into the hole because the rivet cooled from orange to useless in the time it takes a man to walk ten feet at an ordinary pace. And on a wooden plank four hundred and thirty feet above the bay, nothing was done at an ordinary pace. The tower the gang was riveting still stands. It rises five hundred and twenty-six feet above the water line.
You can see it from both shores on a clear day. On that morning in 1935, when the worker and his gang were working in the upper parts of that tower, the day was not clear. The tower was inside the fog. The bay was not visible.
The San Francisco skyline to the west was not visible. The East Bay to the east was not visible. The man was performing skilled productive labor at four hundred and thirty feet in the air between two blank white surfaces. This is not an observation made to dramatize the difficulty of the work.
It is a description of what the conditions were. The conditions were what they were. The man was riveting because riveting in those conditions was the job, and the job was the income, and income in the spring of 1935, in the state of California, was what a man with a rivet hammer and a work gang and a position at four hundred and thirty feet above San Francisco Bay had, which was something the men in the unemployment lines in Oakland and San Francisco did not have. The San Francisco–Oakland Bay Bridge is four and a half miles long.
It crosses the bay in two sections separated by Yerba Buena Island, a rocky outcropping in the middle of the bay that engineers used as a pivot point for a tunnel connecting the western and eastern halves of the crossing. The western half is a suspension bridge with two main towers and a central anchorage. The eastern half is a cantilever truss, a different structural system suited to the shallower conditions of the east bay. The tunnel through Yerba Buena Island that connects them is seventy-six feet wide and fifty-eight feet high, making it at the time of its completion the largest bored tunnel in the world.
At its opening on November 12, 1936, it was the longest steel bridge in the world. It cost seventy-seven million dollars. It took three and a half years to build. Twenty-four men died during its construction, which is the number recorded in the official accident records of the California Department of Public Works.
The number recorded in the official accident records is the number that the official accident records contain. What the official accident records of a Depression-era infrastructure project between 1933 and 1936 contain, and what a complete accounting of a project of this size and duration would require, are two questions with different answers. Eight thousand men worked on it at various stages during construction. At peak employment, the workforce numbered about sixty-five hundred.
These are the figures that appear in the project histories. The men themselves appear in the project histories in aggregate. They appear in the payroll records of the various contractors in detail. The payroll records of the various contractors exist in varying states of availability.
To understand why California spent the years between 1929 and 1933 debating whether and how to build a fixed crossing of San Francisco Bay, you have to understand the cost of crossing the bay before the bridge existed, and what the alternative was, and how inadequate that alternative was. San Francisco Bay is not merely a navigational problem. It is a barrier. The bay runs sixty miles from north to south and between three and twelve miles from west to east at various points along its length.
The city of San Francisco sits at the tip of a peninsula on the western shore. The cities of Oakland, Berkeley, Alameda, and Richmond lie in the East Bay on the eastern shore. The distance between the San Francisco Ferry Building and the Oakland waterfront is about four miles across the water. By land, before the bridge, there was no route that did not require going around the southern end of the bay, a distance of forty-five miles that took two hours of driving under 1930s traffic conditions, and longer when the roads were congested, which they were increasingly as California’s car ownership rate rose through the 1920s at a pace the state’s infrastructure could not absorb.
The ferries operated. They had operated since 1850. By the early 1930s, the San Francisco Bay ferry system was the largest ferry operation in the world by passenger numbers. The Key System, Southern Pacific, Northwestern Pacific, and the state ferries together moved fifty million passengers a year across the bay on a schedule refined over decades to meet the demands of commuters moving between the East Bay and San Francisco.
The ferries did their job. The ferries operated in the way ferries operate, meaning they operated within the limits of vessel capacity and trip frequency, and the fact that a passenger arriving at the terminal at the wrong time had to wait for the next sailing. At the traffic volumes of the late twenties and early thirties, the ferry system was running at its absolute maximum. And the maximum of what a ferry system can deliver is not necessarily adequate.
A proposal for a fixed crossing of the bay had been studied again and again since the 1870s. Engineers studied the problem and returned with the same conclusions each time. The bay was wide, the currents were strong, the fog was persistent, the bedrock under the western bay lay deep beneath a thick layer of bay mud, and the cost of a crossing capable of bearing vehicular traffic would be too high relative to the traffic volumes such a crossing could generate. The studies were completed and then filed away.
The ferries kept running. What changed between 1929 and 1933 was not the bay. It was not the bedrock. It was not the currents or the fog.
What changed was the Depression, and what the Depression did to the political economy of public works financing in California and Washington. The state of California did not have the credit capacity in 1931 to finance a bridge at the cost the engineers projected without assistance. The bond markets in 1931 were not prepared to lend at acceptable interest rates to a state whose public revenues were shrinking as the economy contracted. The state had a political base on both shores of the bay that wanted the crossing built, and an engineering problem that was solvable given sufficient money.
The money was available in Washington through the Reconstruction Finance Corporation, a federal agency created in 1932 to extend credit to states and public bodies for infrastructure projects that private markets would not finance under Depression conditions. The Reconstruction Finance Corporation lent the state of California sixty-one million dollars for the bay bridge in 1933. The state issued bonds to cover the remainder. The Bay Bridge Toll Authority was established to issue the bonds, collect the tolls, and retire the debt.
This mechanism was the same one that would be used to build toll infrastructure throughout the United States over the next two decades. A public authority borrows against the projected toll revenue of a facility that does not yet exist, builds it with the proceeds, and repays the debt from the tolls the facility generates once it operates. For the mechanism to work, the engineers had to prove the reliability of the design, and the construction contractor had to prove the reliability of the schedule. Both conditions were met, which is why the bridge exists today.
The chief engineer on the project was Charles H. Purcell, director of public works for the state of California. Purcell had built roads and bridges across California for two decades before the bay bridge commission came to him. He had never built anything of this scale.
No one in the United States had built anything of this scale. The engineering team Purcell assembled through 1931 and 1932 worked on characterizing the problem, the only way to design something that must stand in a particular body of water under particular conditions for the service life of the facility. They sounded the bay floor at intervals across the route of the proposed crossing and took samples of the bay mud, sending the samples to the laboratory to determine what lay beneath the mud and at what depth it could be found. What they found in the western bay, where the main suspension towers would have to stand, was bedrock at depths of one hundred and sixty to two hundred and twenty feet below the surface of the bay, beneath a layer of mud of varying thickness that could not support a structural foundation without reaching rock.
What they found in the eastern bay, where the cantilever bridge would stand, was shallower water, less deep mud, and firmer materials at more accessible depths. The western crossing was the real engineering problem. The eastern crossing mattered. The western crossing was the thing that had stopped every previous proposal.
When the engineers came to the question of how to found a tower in water one hundred feet deep above mud overlying bedrock at two hundred and twenty feet below the surface, the answer Purcell’s engineers produced was the open caisson. The open caisson is a large hollow concrete structure, open at the bottom and closed at the top, built on the surface, then towed into position and sunk by adding weight while material is removed from within. Workers operate inside the enclosed lower chamber of the caisson, using compressed air to keep bay water from flooding the work space while the caisson descends through the mud toward bedrock. When the caisson reaches the founding stratum the engineers specified, concrete is poured to fill the interior, and the caisson becomes the foundation on which the pier and tower are built.
The compressed air that makes caisson work possible at depth is also what makes it dangerous in a way every engineer understood in 1933, and every worker who entered the compressed chamber experienced directly. At the pressures required to keep water out of the working chamber at one hundred feet below the surface of the bay, the nitrogen in the air the workers breathe dissolves into their blood and body tissues under that elevated atmospheric pressure. If the worker leaves the compressed chamber too quickly, the dissolved nitrogen comes out of solution, forming bubbles in the bloodstream and tissues, causing a condition called decompression sickness, whose symptoms range from severe joint pain that doubles the body over, to paralysis or death. The name the workers gave this condition was the bends.
The engineering profession had been aware of the danger of the bends since the caisson workers who founded the Brooklyn Bridge piers in the 1870s surfaced from their compressed-air chambers unable to walk. The protocol for dealing with decompression sickness was gradual decompression in a recompression chamber, followed by a gradual return to surface pressure. The recompression chambers for the bay bridge caisson work were on site. The protocol was known.
The protocol was followed with the consistency that major construction projects operating on a tight schedule under 1933 conditions follow it, meaning it was applied until it stopped being applied, and the accident records contain what those failures contain. Work on the caissons for the main piers of the western suspension bridge began in 1933. The two main towers of the western bridge were to stand in bay water sixty to one hundred feet deep above bay mud, which in its deepest parts extended one hundred feet further before the excavation operations found material suitable for founding. The caissons for these piers were the largest concrete caissons built up to that time in the United States.
Each was sunk to the depths determined by the soundings by workers operating under compressed air, removing bay mud from within the structure to allow it to descend under its own weight and added weight as construction progressed. Currents in the western bay channel reach three or four knots during strong tidal periods. A four-knot current moves water faster than a man can comfortably swim in calm conditions. The workers building the cofferdams around the caisson sites to deflect the current during construction worked from barges that the current constantly pushed against the protective structure.
The submerged elements of the bay bridge were built in the same bay used by ferries and commercial shipping and strong tidal currents. And the fog that crossed the Golden Gate on its own schedule, indifferent to the construction schedule, arrived and settled, while work continued or stopped as conditions permitted. The towers rose in 1934 and 1935. The American Bridge Company fabricated the steel for the towers of the western suspension bridge and delivered it to the site by seagoing barge.
Each section of tower steel was lifted by crane above the section beneath it, and the riveting gangs followed the crane to drive the rivets connecting each new section to the structure below. There were approximately one million three hundred thousand rivets in the Bay Bridge. Each rivet was driven by a four-man gang working with a portable furnace and a pneumatic hammer, in a rhythm four men develop when they perform the same work together for months. The riveting gang is the structural unit of the steel bridge construction workforce.
The heater, the catcher, the bucker-up, and the riveter are four roles that only function together. A fast heater working with a slow riveter produces cold rivets and wasted work. A slow heater working with a fast riveter produces the same result. The gang works at the pace of its least efficient member, and the least efficient member on a good production day produces rivets that will hold structural steel for the service life of the structure, which in the case of the Bay Bridge has now exceeded eighty-eight years and continues.
The men who drove those rivets came primarily from the skilled metal trades that built the steel infrastructure of the American industrial economy. They were mostly white, mostly organized under the International Association of Bridge, Structural, Ornamental and Reinforcing Iron Workers, and paid wages reflecting their trade classification, about one dollar twenty to one dollar fifty an hour in 1934, at a time when the average manufacturing wage in California was far lower, and the men standing in the unemployment relief lines in Oakland earned nothing at all. The unskilled labor on the Bay Bridge, handling materials and operating equipment under supervision, doing the physical work that required no journeyman classification in a recognized trade, was drawn heavily from the flood of men who came to California in the Depression years, from the rural South, from Mexico, from the Philippines, and from the Central Valley agricultural labor market that collapsed along with everything else in 1929 and never recovered. These men were paid the common labor rate, which in 1934 was about sixty-five cents an hour.
The difference between a dollar fifty and sixty-five cents an hour is not a small difference in a Depression-era household budget. The division between skilled trades and common labor on the Bay Bridge followed the racial lines that governed employment in California construction in the 1930s, with the same consistency and for the same reasons those lines governed employment throughout the industry in those years. Equipment operator jobs and supervisory positions were held mostly by white men. Common labor was performed largely by Black men and Mexican and Filipino workers, who occupied the lowest rungs of the California labor market in the years before the war and defense-industry employment would rearrange some of those rungs temporarily and incompletely.
Black workers on the Bay Bridge construction do not appear regularly in the project histories written in the decades afterward. The project histories were written from the engineering records and official accounts of the state agencies that administered the project, and those records documented the structure in great detail, the workforce in aggregate, and individual workers in the payroll books. The payroll books are where the names are. The project histories are where the structure is described.
Men applied for jobs on the bridge project in numbers exceeding available jobs by ratios that varied by construction phase and trade classification, but were never small. A job on the Bay Bridge in 1934 was a job in an economy where employment opportunities were scarce enough that men traveled long distances to apply for them, and waited in lines for the chance to wait in longer lines. The wages paid reflected the conditions of that market, in the sense that they reflected the relative power of the men setting wages versus the men receiving them. And the relative power in 1934 in the California construction industry was not ambiguous.
The fog that came through the Golden Gate in the morning did not come on a schedule. It came with the weather systems and the temperature differentials between the cold Pacific waters and the warmer inland valleys. The fog settled over the bay surface and reduced visibility to the distance between one man and the next on the same scaffold plank. Ironworkers at four hundred feet in dense fog worked inside a white envelope from which the surface of the bay below was not visible.
The shore in any direction was not visible. The visible world was the steel they worked on, the rest of their gang, and the fog itself. The wind coming off the bay was the other factor. The bay acts as a funnel for marine air crossing through the Golden Gate, and the gap in the coastal mountains at the Gate accelerates the air movement during the afternoon period.
An ironworker on a scaffold four hundred feet above the bay in afternoon winds performed precision work with a pneumatic hammer in conditions that pushed him sideways against the scaffolding. When the wind exceeded the threshold for safe work, work stopped. The construction records logged the weather delays. The bond obligations of the Bay Bridge Toll Authority contained no weather clause.
The construction sequence on the western suspension bridge moved from foundations to towers to cables. The caissons were sunk first. The piers were built atop the caissons. The towers were erected above the piers.
Then cable spinning began. Cable spinning on a suspension bridge is the process by which individual wires are carried back and forth across the bridge on a traveling wheel. With each pass of the wheel, an additional wire is laid alongside the wires already in place. The accumulated wires are eventually compacted to form the main cables that carry the weight of the bridge below.
Each main cable on the western half of the Bay Bridge contains seventeen thousand four hundred and sixty-four individual wires. The wire was carried across the bridge on traveling wheels that covered the full distance between anchorages on each pass. The men who ran the spinning operation followed the wire across heights and distances that required them to work continuously. On the eastern half of the bridge, the cantilever truss was built in a different sequence, but with the same basic materials: structural steel, rivets, and the work of men working at height above the bay in the conditions the bay imposed.
A cantilever grows from anchor arms toward the center of each span, with the structure extending from each pier toward the open gap it would eventually close when the two cantilever halves were joined at the suspended span in the middle. The ironworkers followed the advancing steel, fastening each new section to the structure behind it, with the work front moving toward the open water ahead of them. The bridge deck was built in a two-level design. The upper deck was for vehicular traffic.
The lower deck was for the Key System electric trains and truck traffic. The two-deck design was the engineering solution to the traffic demand the bridge was designed for, which included not only toll-paying automobile commuters but transit passengers whose trains needed to cross as much as cars did. Work proceeded on both decks from both ends toward the center simultaneously, with the construction crews meeting in the middle of the bay to close the two halves of the structure that both shores had contributed to from opposite directions for three years. The Yerba Buena Island tunnel, connecting the western and eastern halves of the bridge, was excavated through the island’s solid rock by blasting gangs working from both ends of the tunnel alignment at once.
They drilled holes into the rock face, filled them with explosives, detonated the charges, and removed the broken rock before repeating the process. The dust in the tunnel after a blast and before ventilation was the dust the workers inhaled because the nature of the work required it. Ventilation cleared the air to the level it was designed for, which was not the level of clean air, just as working conditions inside a 1934 tunnel were not comparable to construction sites on the surface. On November 12, 1936, the Bay Bridge opened to traffic.
California Governor Frank Merriam drove the first car across the bridge from the San Francisco toll plaza to Oakland in a procession of official vehicles. President Roosevelt in Washington pressed a telegraph key that sent a signal over Western Union wires to a relay on the bridge, opening the toll gates at both ends at the same moment. Speeches were given at both ends. The opening photographs show the governor, commission officials, the senior engineers, and the politicians who had championed the bridge project through the years that preceded its construction.
The men who drove the one million three hundred thousand rivets do not appear in the opening photographs. The riveting gangs that completed the structure above San Francisco Bay in the fog of 1934 and 1935 had moved on to new job contracts by the time the governor crossed the bridge in November 1936. The specialized metal construction workforce of California and the West Coast did not stay at a job site after the contract ended. The Bay Bridge work finished, and the ironworkers moved to the next job, whether a bridge or a building or another project starting up somewhere else.
The officials who appear in the opening photographs are in the administrative records of the bridge toll authority. The engineers are in the technical literature. Governor Merriam is in the political history of California. The men who excavated the Yerba Buena tunnel, dove into the caissons, erected the towers, spun the cables, and drove the 1.
3 million rivets are in the payroll records of the American Bridge Company and the other subcontractors, in whatever form those records survived. The toll at opening was sixty-five cents per car, collected on the San Francisco side in both directions. The ferry operators on the bay watched their passenger numbers collapse in the months after the opening, as drivers who had been accustomed to using the ferries realized that the fixed crossing was more convenient, faster, and more predictable, and chose accordingly. The Key System and Southern Pacific ferries reduced service.
They eventually stopped. The ferries that had carried fifty million passengers a year across the bay before the bridge continued for a few more years on reduced schedules, then ceased operation. In 1989, the Loma Prieta earthquake struck the San Francisco Bay Area at 5:04 p. m.
on October 17 during the evening rush hour. The earthquake caused a section of the upper deck of the eastern cantilever bridge to collapse onto the lower deck. One driver was killed. The bridge was closed and not reopened for a month.
The 1936 structure had survived fifty-three years in a seismically active environment. It had been designed to withstand, according to the engineering knowledge of 1933, which was not the same engineering knowledge as 1989. The caissons that workers had sunk into the bedrock in 1933 and 1934 under compressed air in the strong current of the bay’s western channel held. The steel that the riveting gangs had joined piece by piece in the fog of 1934 and 1935 held in the sections not affected by the deck collapse.
The eastern bridge was eventually replaced. The new eastern span opened in 2013, a single-tower self-anchored suspension bridge designed to current understanding of the bay area’s seismic risk. The original 1936 cantilever truss was demolished. The western suspension bridge, the half of the bridge where the riveting gangs had worked in the fog in the spring of 1935, retains its original structure.
The towers are the towers that still stand in the bay today. They have been repainted several times. They have been inspected continuously. The rivets are the rivets those gangs drove in 1934 and 1935.
Near the middle of the bridge, on a clear day, you can see both shores. The San Francisco skyline lies to the west. The East Bay hills lie to the east. On a foggy morning, neither is visible.
The bridge deck is visible for a certain distance in each direction, then the fog closes in. These are the conditions in which the riveting gangs worked every foggy morning for the three and a half years of construction. They could not see the shore they had left. They could not see the shore they were building toward.
The survey marks on the steel located them on the structure. The rivet hole in front of them was the reference point. They drove the rivet, moved to the next hole, drove that one. If you drive across the Bay Bridge on a foggy morning at the speed the structure allows, you pass the location where those gangs worked in about four seconds.
The tower goes by the window. The cable runs overhead. The bay lies one hundred feet below you, and on a foggy morning it cannot be seen. The structure between you and it is the structure those men built.
The rivets they drove in the conditions the bay created, in the years when the Depression was reshaping the American labor market, while the money to build the bridge came from Washington because the bond market would not supply it on acceptable terms without federal backing. The bridge is still there. The western suspension bridge is the 1936 structure, eighty-eight years old, still carrying the traffic the bay generates daily between the city and the East Bay. The fog comes through the Golden Gate on its own schedule.
The tidal current runs at four knots through the main channel at strong tide. The caissons the workers sank into the rock in 1933 and 1934 under compressed air in moving water remain on the bay floor where they were placed. The twenty-four names in the official accident record are the twenty-four names the record contains. The names of the eight thousand men who worked on the project at various stages over three and a half years are in the contractors’ payroll records in whatever form those records survived.
Their names are not on any plaque at the San Francisco toll plaza. The plaque carries the names of the commissioners, the engineers, the governor, and the officials of the agencies that authorized and financed the crossing. The tower is still there. The rivets are in it.
The bridge carries the traffic you drive.
