In the autumn of 1932, a carpenter named Refugio Méndez stood on a wooden platform 160 feet above the floor of Carrizo Gorge. Through the open framework of a bridge that was not yet finished and would not be for another four months, he felt the valley winds rising from the desert bottom below, strong enough to push his body sideways on the platform without warning. He was working in the Anza-Borrego desert of Southern California, in a gorge that the San Diego and Arizona Eastern Railway had been trying to cross with trains for twenty years. The gorge is 2,000 feet deep at its widest point and runs eleven miles through the Peninsular mountain ranges, between the floor of the Imperial Valley and the coastal plain at San Diego.
The railroad had already crossed the gorge once, in 1919, over a wooden trestle that took two years to build. But it burned completely in the afternoon of a single day in 1926, six months after it opened, when a spark from a passing locomotive fell onto the wooden structure, and the desert air finished the job faster than the fire crew could reach it from the nearest station. The railroad needed to rebuild. It had already rebuilt once, in a temporary form, and that temporary structure was nearing collapse.
What Refugio Méndez was doing, 160 feet above the canyon floor in the autumn of 1932, was building the permanent replacement: a curved wooden trestle, 787 feet long and 185 feet high at its tallest point, the equivalent of an eighteen-story building measured from its base to the rail it carries. It was being erected over a desert gorge in the second year of the Great Depression by men who were paid in company scrip, lived in tents, and had no practical alternative but to accept it, because the alternative was the Imperial Valley labor market of 1932, which was worse. The bridge he was building is still there. It is called the Goat Canyon Trestle.
According to everyone who has measured it, it is the tallest remaining curved wooden railroad trestle in the United States, and by most accounts, the tallest curved wooden railroad trestle ever built. You cannot reach it by car. You cannot see it from any road. The railroad that built it went bankrupt in 1933, a year after the bridge was completed, and the line through Carrizo Gorge has been out of regular service since 1976.
To reach the wooden trestle today, you must walk fourteen miles through the Anza-Borrego Desert State Park, along an unmaintained route that requires a permit and a level of desert navigation that most people who want to see the bridge discover they lack only after they are already several miles from the start. The men who built it have no memorial at the trailhead. The wooden trestle is what has the memorial at the trailhead. Those men exist in the employment records kept by the San Diego and Arizona Eastern Railway in 1932, and in whatever form those records survived the bankruptcy proceedings of 1933, the subsequent reorganization of the company, and the ninety-two years since the last trestle was driven into place.
This is the story of how they built it, what it cost, and why that cost does not appear in the official account. To understand why a railroad needed to cross Carrizo Gorge at all, you have to understand what the gorge was and what lay on both sides of it. San Diego in the early twentieth century was a coastal city with a harbor and a naval facility, growing at a rate that the existing rail lines could not serve efficiently. The Southern Pacific Railroad, which controlled the main line between San Diego and the rest of the country, routed its traffic north through Los Angeles.
Any shipper in San Diego who wanted to move goods east across the Southwest had to send them north to Los Angeles first, then east through the San Bernardino Pass. That added hundreds of miles and a surcharge imposed by the Southern Pacific monopoly on every ton that passed through San Diego. John D. Spreckels, who owned the San Diego Union-Tribune and much of the commercial real estate in downtown San Diego, and who had a financial stake in San Diego’s competitive standing that made money a personal matter, saw that the answer to the Southern Pacific monopoly was a competing rail line running directly east from San Diego across the mountains to the Imperial Valley, where it could connect with the Southern Pacific’s transcontinental routes at Yuma and give San Diego shippers a direct path east that did not pass through Los Angeles at all.
The engineering problem lay in the mountains between San Diego and the Imperial Valley. The Peninsular ranges run north to south along the eastern edge of the San Diego coastal plain, rising to between 5,000 and 6,000 feet before dropping abruptly to the floor of the Imperial Valley at elevations below sea level. The descent from the top of the range to the valley floor occurs over a linear distance of about thirty miles and a change in elevation of 6,000 feet. A railroad grade required either a very long route to reduce the steepness of the slope, or a direct route with grades steep enough to challenge the locomotives of the early twentieth century.
The most direct route through the mountains, which Spreckels’s engineers identified as practical, passed through Carrizo Gorge, a canyon carved by Carrizo Creek through the eastern face of the ranges over millions of years of intermittent flooding. The gorge was narrow and steep-walled, full of blind turns that put the rail route in direct confrontation with the canyon walls at several points. The engineers counted seventeen tunnels required in the gorge section alone. They considered the Goat Canyon bridge a separate problem of a different kind, because a tunnel is a hole in rock, and rock is a comprehensible material; a bridge rising 185 feet above the floor of a desert canyon was a dilemma in wooden engineering over a bottomless void, worth considering from the perspective of a man who might fall into it.
Construction of the San Diego and Arizona Eastern Railway began in 1907 and continued through the intermittent financial difficulties that characterized most American railroad construction projects of that period. Spreckels had enough money to start the line. He did not have unlimited money to finish it on a schedule dictated by engineering complexity rather than cash flow, and Carrizo Gorge was the most complex section of the hardest stretch of the railroad. The tunnels alone required years of drilling and blasting through rock that geologists described as hard and that the diggers found harder.
The Carrizo Gorge section was the last section completed on the railroad. The line reached Yuma and its connection with the Southern Pacific in 1919, twelve years after construction began, at a total cost that far exceeded the estimates the engineers had given Spreckels when he committed to the project in the first years of the century. The first Goat Canyon trestle was completed in 1919 as part of the final push to close the last gap in the Carrizo Gorge section. It was built from Douglas fir timbers, the standard material for wooden railroad bridges in the American West in the early twentieth century, because that timber offered the required sizes and strength in compression, and because it was cut from the forests of the Pacific Northwest and shipped by sea to San Diego in the quantities the construction demanded.
The first bridge lasted six years. In 1925, a spark from a locomotive ignited the wooden timbers. The fire spread through the dry desert air faster than wooden structures burn in wetter climates, and the bridge was gone before the fire crew could arrive. The railroad rebuilt.
A second bridge was rushed into place to restore service across the gorge. It was built with the speed that emergency work requires and without the full engineering review that any permanent structure would have demanded. It served its purpose. It was not built to last.
By 1931, the second bridge was deteriorating at a rate that the railroad’s maintenance staff assessed as unacceptable for continued service under freight train loads, and the railroad’s engineering department began designing a permanent replacement. The engineering department of the San Diego and Arizona Eastern Railway in 1931 was not a large organization. Neither was the railroad itself. Its track ran 148 miles.
Its traffic depended on the agricultural and industrial freight of the Imperial Valley and the commercial traffic of the Port of San Diego. It was not profitable in the way transcontinental railroads calculated their earnings. But it was necessary in the sense that John Spreckels had understood when he committed to building it. It gave San Diego a rail outlet that did not pass through the Southern Pacific hub at Los Angeles.
The existence of that outlet had material value to San Diego shippers regardless of whether the railroad returned a profit on the capital invested in building it. The design of the permanent replacement for the Goat Canyon trestle followed the engineering logic that had governed the original. Douglas fir structures, trestle framing, and the same curve alignment the gorge required, with a rail height of 185 feet above the canyon floor at the highest point. The curve was not optional.
The walls of the gorge forced the track to take a curve of approximately ten degrees as the railroad threaded its way between the western canyon wall and the narrow terrace above the creek. Building a straight bridge at that location would have required either cutting into the canyon wall from the west side or extending the bridge south to a point where the gorge was wide enough for a straight crossing. Neither option was economically justified. So the curve was built into the design from the first drawing.
Building a curved wooden bridge is harder than building a straight one for reasons that may not be immediately obvious to someone who has never built wooden bridges. A straight bridge relies on repetition. The trestle bents, the vertical framing units that carry the deck loads down to the foundations, are identical from end to end except where the ground elevation changes. The joints are identical.
The lengths of the pieces are equal. A crew that builds the first bent correctly has a template for every subsequent bent. A curved bridge does not rely on that repetition. Each bent sits on a radius slightly different from the one before it.
The cap, the horizontal beam at the top of each bent that carries the stringers and rails, must be cut at an angle matching the curvature at that specific point in the track. The stringers spanning between bents must account for the chord distance across the curve in each section. The joints must be designed to work at the angle the curve imposes at each location. A carpenter who cuts the wrong angle into a cap timber has wasted the effort of the cut and the cost of the lumber and must start the piece over.
At a height of 185 feet, cutting a piece of timber to the wrong dimensions, lowering it to the ground, and bringing the correct piece back up to the work platform is not easy. The timber for the Goat Canyon trestle was Douglas fir, as it had been for the original. It was shipped by rail from the Pacific Northwest to San Diego, then brought by a narrow-gauge work train into the gorge along the existing track south of the bridge site. The dimensions were what wooden bridge engineering of the period demanded: 12×12 and 14×14 inch cap timbers, 6×8 and 8×10 stringers, 10×10 and 12×12 posts for the trestle bents, and heavy timber sills for the foundation bents wherever the rock allowed direct loading.
The bolts were iron drift pins driven through pre-drilled holes, and the connections were made with the iron hardware specified by the engineering drawings. Getting those materials to the bottom of Carrizo Gorge and then lifting them to the required height for the framing was the first and fundamental problem of the construction, and it was a problem made far harder by the gorge’s isolation than it would have been at a site with road access. The men who built the Goat Canyon trestle in 1932 came from the labor pool that the Imperial Valley in San Diego County provided in the second year of the Great Depression. They were mostly Mexican and Mexican-American workers.
The Imperial Valley in the early 1930s had one of the largest Mexican agricultural labor populations in the United States, a workforce that had come north during the decades of the Mexican Revolution and the agricultural migration that followed, to fill the demand for field labor in the irrigated agriculture of the Imperial Valley, which grew cotton, melons, and lettuce and required large numbers of workers at harvest time, paying wages set by the growers and accepted by the workers because the alternative was returning to a Mexico that offered less. Work on the railroad paid better than work in the fields. It was harder, more dangerous, and intermittent, but the daily wage for a laborer on a construction project in 1932 was higher than the wage for an agricultural worker, even in a year when construction wages were falling with the rest of the economy. The men who applied to work on the replacement of the Goat Canyon trestle in the summer of 1932 were people for whom the construction wage, whatever it was that year, was an improvement over what the valley’s agricultural economy was offering.
The railroad paid in the currency that the Depression economy had made common among employers who understood that their workers had no practical choice, which was company scrip redeemable at the company store instead of the cash wages that the railroad’s treasury, strained by the shrinking freight market of 1932, preferred to defer paying. The scrip was legal. It was common. It was an arrangement that returned part of the wage to the employer through the profit margin on goods that could be bought with scrip at the company store compared to the cost of the same goods at a general store in El Centro or Brawley where cash was accepted.
Whether the men who built the Goat Canyon trestle understood the scrip arrangement before they arrived at the work camp in Carrizo Gorge or discovered it upon arriving is a question the employment records might answer if they still exist. What the arrangement meant in practical terms was that a man who spent three months in the gorge building a bridge 160 feet above the canyon floor in the heat of a California summer and the winds of its autumn left the work with less purchasing power than his nominal wage indicated. The work camp was inside the gorge. The gorge provided no permanent settlement of any kind.
The camp was tents on the canyon floor near the bridge site, with cooking facilities sufficient to feed a crew of the size the construction required, and water brought by tank cars on the work train from the nearest reliable source outside the gorge. The Imperial Valley summer of 1932 was what it always is. The heat exceeds 100 degrees Fahrenheit in the shade on most days between June and September, with shade being a limited resource on the floor of a desert canyon where the walls block direct sun for parts of the day but do nothing to reduce the radiant heat that the rocks absorb and re-emit onto the men living among them all night. The men working on the high platforms of the bridge under construction were not in the canyon floor shade.
They were in full sun for the entire workday when the sun was above the canyon walls, which covered most of the working hours in summer. A platform 100 feet above the canyon floor, on the west-facing side of a bridge in the Anza-Borrego desert in July, is a platform with no shade and no air movement, except for the thermal winds generated by the heated canyon walls in the afternoon, which arrive with enough force to push a man who does not expect them. The tools were iron and wood. The iron components of the trestle hardware were hot enough in the direct August sun to raise blisters on the hands of anyone who touched them without gloves.
The men wore gloves. The gloves were cloth and leather, absorbed the sweat from the hands inside them, then dried and stiffened and cracked, and were replaced when the railroad’s supply allowed. The risk of falling at the heights that Goat Canyon trestle construction required was the condition that made every other circumstance secondary. A man working on a wooden platform 130 feet above the floor of a rocky canyon is working at a height from which a fall cannot be survived under any normal circumstances.
The canyon floor was not water, as it was for the pile-driving barges on Lake Pontchartrain where a fall was a manageable situation if rescue came quickly. The canyon floor was rock and the dry bed of Carrizo Creek, which carried no water at all in the summer of 1932. The safety technology available to a construction worker in 1932 was a rope tied to a timber. There were no safety harnesses of the design that OSHA standards would later require.
There were no fall-arrest systems. There were ropes. Whether a man tied himself off depended on whether the task he was performing allowed it, and whether the crew foreman’s standard included that requirement. The foreman’s standard reflected what the industry considered normal in 1932, which was that men worked at heights on wooden platforms and sometimes fell, and the industry accepted that as part of the working conditions.
Records of falls at the Goat Canyon trestle construction site do not appear in any publicly available source. The railroad went bankrupt in 1933. The bankruptcy reorganization produced a new legal entity and a new set of corporate records. What survived of the original construction records through the reorganization and the subsequent ownership changes in the following decades is not established in the published accounts.
The names of the men who fell, if any man fell, are not in those accounts. The names of the men who did not fall are not in those accounts either. The bridge is in those accounts. The bridge is what those accounts were written to document.
The sequence of construction of the new Goat Canyon trestle began with the foundation bents at the canyon walls from both the north and south ends, working toward the middle where the height was greatest and the complications of the curve required the most effort. A foundation bent is the lowest structural unit of the trestle, the unit that transfers the load of the bent above it into the ground. At the canyon walls, the rock was close enough to the surface that the foundation sills could be set directly on the rock face with minimal excavation. In the middle of the gorge, where the canyon floor was softer sediment deposited over the bedrock over the centuries since the last major flood had scoured the channel, the foundations required more preparation.
The soil samples taken by the engineers before the final design was drawn showed that the canyon floor material was suitable for the loads the bridge would place on it, provided the footings were sized according to the bearing capacity of the material. The footings were sized accordingly. The construction of the bents used the same system that the railroad bridge industry had developed over half a century of building bridges across the American West. A hand-operated wooden crane at the top of the rising structure, powered by horses on the canyon floor below, lifted the framing components through a system of pulleys from the supply pile to the height where the carpenters were working.
The wooden crane was the technology. The horses were the power source. A team of four horses operating the crane in 1932 could lift about 1,500 pounds to a height of 150 feet at a rate determined by the engineering of the pulley system. A 14×14 inch timber sill thirty feet long weighs about 1,800 pounds.
Moving it from the supply pile on the canyon floor to the framing platform 130 feet up required two lifts by the horse-powered crane at the bridge’s average height, and required the carpenters on the platform to guide the timber into position while it swung on the crane hook in whatever wind was blowing through the gorge at that moment. The gorge winds were not predictable. They came from the south in the morning as the desert heated and the air rose, drawing cooler air from the coast through the lower part of the gorge, then came from the north in the afternoon as the thermal pattern reversed. The winds moved an 1,800-pound timber hanging from the crane hook.
The men on the platform guided it with wooden poles and their hands, using the experience they had gained on previous bridge projects. When the wind blew hard enough that guiding the timber was beyond skill and became a matter of luck, the crane was stopped, the timber was lowered, and the crew waited. The waiting was not paid. The construction contract, as was customary for contracts of that period, specified payment for work performed, and weather delays were conditions borne by the contractor, not the railroad.
A man who waited two hours in the valley heat for the afternoon thermal winds to ease before the crane could be restarted and the timber set in place was a man who spent two hours in the sun on a wooden platform above a rocky gorge, and was not paid for the waiting nor for the exposure. The mortar used for the stone foundation footings was hydraulic cement, just as it was for the Starrucca bridge eighty years before, and as it remains for any construction work in an environment adjacent to water where ordinary lime mortar cannot dry reliably. The cement was brought into the gorge by the work train and mixed at the footing sites with water brought from the tank car. The mixing was done by hand, as it was at remote construction sites in 1932, in wooden mixing boxes with iron shovels by men classified on the payroll as common laborers, and their wages reflected that classification.
The curved deck of the wooden trestle required stringers and cap timbers cut to angles that the carpenter had to transfer from the engineering drawings to the lumber itself with precision, without the benefit of the digital tools that later generations of builders would use for the same process. The drawings showed the angle. The carpenter marked the angle on the face of the timber using a framing square and a pencil line, then cut it with a hand saw or a two-man crosscut saw, depending on the dimensions and the workspace available at the site. Cutting a 14×14 inch Douglas fir cap timber with a two-man saw took several minutes of sustained effort by two experienced men.
If the cut was made at the wrong angle, those minutes were lost and so was the timber. At a height of up to 185 feet at the highest point, the consequences of cutting a timber wrong were multiplied by the cost of transporting any replacement to the work site. The supply chain brought materials to the canyon floor. The horse-powered crane lifted materials from the canyon floor to the framing platform.
A piece of timber that had to be lowered and brought back up again after correction had consumed the labor of the horses, the work of the carpenters, and the time of a working day that the construction schedule had not accounted for, a schedule the railroad’s engineering management was tracking against the deterioration of the temporary bridge the permanent structure was replacing. The engineering supervision on site was provided by the railroad’s own engineering staff, under the chief engineer of the San Diego and Arizona Eastern Railway. Whether the level of on-site supervision was continuous or periodic is not established in the accounts available to the public. The construction was performed by a contractor, and the division of responsibility between the contractor’s superintendent and the railroad’s engineering staff for daily construction quality was the division specified in the contract.
What the contract specified has not been published in a form readily accessible to historical researchers. What the completed bridge showed when the permanent structure was finished and the work train crossed it with its first load in late 1932 was that the angles had been cut correctly, that the curve had been built to the radius the drawings specified, and that the rails ran smoothly across 787 feet of wooden framing without the alignment errors that would have resulted from a wrong cut in the cap timber or a post out of plumb. The structure carried the working loads the railroad imposed on it. The curve was correct.
The framing was sound. The men who made the framing sound are not in the photographs taken at completion. The completion photographs, if they exist and if they have been preserved in a form accessible to researchers, show the bridge. Whether those photographs show the men who built it, and whether those workers were of the type that supervisors and railroad management considered appropriate to document in completion photographs, is a question about 1932, and the answer of that year is evident in the systematic absence of Mexican and Mexican-American construction workers from the official documentary records of infrastructure projects in the Southwest during that decade.
The San Diego and Arizona Eastern Railway went bankrupt in November 1933, thirteen months after the Goat Canyon trestle was completed. The bankruptcy was not caused by the bridge. The bankruptcy was caused by the collapse of freight traffic volume and freight rates in the Great Depression, which made the economic viability of operating a short 148-mile desert railroad unsustainable against the debt service on the capital the railroad had consumed to build it. The bridge faced no such crisis.
The bridge was complete. It was carrying train loads. It was doing what it was built to do and would continue to do regardless of what the railroad’s financial condition was going through in the courts. The bankruptcy reorganization produced a reorganized company that continued to operate the line.
The line through Carrizo Gorge remained in service under various ownerships and reorganizations through the middle decades of the twentieth century. The agricultural and industrial freight of the Imperial Valley continued to move east through San Diego on the route that John Spreckels had committed to building in the first years of the century, and that the men in Carrizo Gorge had made possible by drilling seventeen tunnels through the rock and raising a curved wooden bridge 185 feet above the canyon floor. In 1976, a flood in Carrizo Creek caused heavy damage to portions of the railroad through the gorge. The cost of repair exceeded what the line’s owner at that time considered justified by the traffic revenue.
Service on the line through Carrizo Gorge was discontinued. It has not returned to regular service since. The trestle was not damaged in the 1976 flood. The trestle stood above the flood level.
The trestle stood while the railroad beneath it ceased operation. The trestle has stood since, through the desert summers and winters, and the flash floods that occasionally sweep the canyon when summer monsoon winds bring enough rain to the watershed above the gorge to produce a flow faster than the canyon floor can absorb. The Anza-Borrego Desert State Park, which contains the land around the trestle, has managed the site since the line went out of service. The trestle lies within the park boundaries.
Access is permitted on foot. The fourteen-mile round trip to the trestle is long, and the desert terrain is difficult enough that the number of people who reach it in a given year is small compared to the number who express interest in reaching it. People who reach it find a wooden trestle 185 feet high and 787 feet long, standing in a desert canyon for ninety-two years without the maintenance a structure in active railroad service would have received. The timber has weathered.
The iron fittings have corroded. The frame remains intact in its structural configuration. The curve is the same curve the engineers drew and the carpenters built in 1932. The desert preserves timber better than a wet climate does.
Rainfall in the Anza-Borrego desert is very low in most years. The timbers produced by the Douglas fir forests of the Pacific Northwest in the 1930s, cut to the dimensions the framing required, and bolted together with iron drift pins by men paid in scrip and housed in tents on the canyon floor, have survived ninety-two years in a climate as harsh on human comfort as it is suited to the long-term preservation of wood. Douglas fir still does what it does when it is dry and protected from the mechanical damage that traffic loads impose. It holds its shape under the compressive load of its own weight.
The joints are the weakest elements at this age. The drift pins corrode. When a drift pin corrodes completely, the joint it held begins to move under wind loads and thermal expansion and contraction. The park manages what it can manage.
The trestle is not a maintained structure in the sense that an active railroad bridge is maintained. It is a historic structure in a remote desert park accessible on foot, managed for preservation rather than operation. If you hike the fourteen miles today to the Goat Canyon Trestle, which requires a permit from the Anza-Borrego Desert State Park and a level of physical fitness the desert does not forgive in those who lack it, you will arrive at a structure visually unlike any wooden construction you have seen in your life. The curve is the first thing you will notice.
A straight wooden trestle of this height is impressive as any eighteen-story wooden structure. A curved wooden trestle of this height is something different, because the curve makes the engineering choice visible and obvious. This structure is here because the canyon walls demanded it, and the men who built it accepted the complications the canyon demanded, incorporating the curve into every cap timber, every brace joint, every bent frame over the 787 feet of the bridge’s length. The second thing is the size of the timber.
The cap timbers are 14×14 inches in cross-section, a size larger than most people who have not worked in heavy timber construction will ever encounter. They are not the standard 2×4 and 2×6 framing lumber of residential construction. They are structural members of a bridge designed to carry loaded freight trains, and their size reflects what that demanding task required. The third thing, which you notice when you have stood at the bridge long enough to stop studying the structure and start studying the place, is that the desert is silent the way remote places are silent when no mechanical equipment is operating within their reach.
The canyon carries the sounds of wind and birds and the occasional movement of the creek when there is water in it. It carries no sound that would have been unfamiliar to the men who built this bridge in 1932. Those men drove heavy iron drift pins through pre-drilled Douglas fir timbers 160 feet above the canyon floor. They did it with crude wooden cranes and pulleys and simple carpenter’s tools and their bare hands.
They did it in the peak heat of the Anza-Borrego desert summer and in the canyon winds that came up from the desert floor in the afternoons without warning, moving the timbers they were trying to guide into place. They were paid in scrip. They lived in tents. Their names are not at the trailhead.
The bridge they built in thirteen months still stands after ninety-two years in a desert canyon, where almost nothing built by human hands in 1932 has survived the intervening decades intact. The foundation material beneath the footings has not moved. The Douglas fir framing they cut to the angles the curve required has never changed its curved shape. The iron drift pins corrode at the rate that iron corrodes in a dry alkaline environment, a rate slow enough that what these men held together, the men whose names are not at the trailhead and the engineers whose names are recorded in the railroad’s design records, remains a structure recognizable as exactly what they built.
The canyon does not know the railroad went bankrupt. The canyon does not know the scrip paid as wages was worth less than its face value at the company store. The canyon knows what it has always known, which is that it is 2,000 feet deep, that the winds blow up from the desert floor in the afternoon, and that the rock walls radiate the heat they have stored all day back into the air above them all night. The wooden trestle is curved because the canyon walls demanded it.
The men who cut the curve into the timber exist in the payroll records of a bankrupt railroad, if those records survived the ninety-two years between the autumn of 1932 and today. The wooden trestle is still there.
