The Iron City Beneath the Old Dam

The dam remained in its altered state. The water pooled behind it through the seasons. The water leaked through it. The seasons turned. The years turned. The dam did not change. The dam stayed as the club had made it — lowered, patched, its spillway narrowed, its discharge pipes removed. Above the dam, the reservoir gathered rain and snowmelt and held them. Below the dam, the valley worked. This was the state of things before the water came in May of 1889, a state that had persisted since the club’s modifications.

In March 1887, the Little Conemaugh rose. The water came down the narrow valley from South Fork and East Conemaugh and pressed into the streets of Johnstown. It filled cellars along Washington Street. It covered the plank flooring of the Cambria Iron Company’s lower rolling mill. It rose against the stone abutments of the Pennsylvania Railroad bridge at the confluence. The men at the furnaces watched the water reach the boiler pads. The men at the rolling mill pulled the hot metal from the rolls and shut the line down. The foreman sent for the pump.

The pump was a stationary steam unit. It sat on a platform above the mill floor. The crew connected hoses. They ran the lines to the cellar doors along Washington Street and to the low passages beneath the rolling mill. The water ran brown. It carried coal dust and cinder and the fine scale of iron working. The pump ran for six hours. The water dropped. The crew swept the mill floor. The foreman ordered the line restarted. The hot metal went back through the rolls. By the afternoon shift, the mill was running at capacity. The cellar held a residue of silt. The silt dried. The foreman noted the shutdown in his log. The entry recorded the cause as high water and the duration as six hours. The entry was one line. The entry was routine.

This was the fourth freshet in two years. The river had risen in the spring of 1885, again in the autumn of that year, and again in the spring of 1886. Each time the pattern repeated. The water came. The water filled the low ground. The pumps ran. The water dropped. The work resumed. The town absorbed the river the way it absorbed the soot from the furnaces — as a cost of where it stood.

The city of Johnstown was founded in 1800 by Swiss immigrant Joseph Johns, anglicized from “Schantz,” where the Stonycreek and Little Conemaugh rivers joined to form the Conemaugh River. It began to prosper with the building of the Pennsylvania Main Line Canal in 1836. Construction of the Pennsylvania Railroad and the Cambria Iron Works in the 1850s brought further industry to town, and eventually led to abandonment of the canal. By 1889, Johnstown’s industries had attracted numerous Welsh and German immigrants to work. With a population of 30, 000, it was a growing industrial community known for the quality of its steel. The high, steep hills of the narrow Conemaugh Valley and the Allegheny Mountains to the east restricted development, keeping it close to the riverfront.

The valley was narrow. The Allegheny Mountains rose on both sides. The Little Conemaugh entered from the east, winding through fourteen miles of restricted gorge. The Stonycreek entered from the south, cutting through a valley only slightly wider. The two streams met at the point where Johns had laid out his town. Below the confluence, the combined river turned west and cut through the mountains toward Pittsburgh. The valley floor at Johnstown was flat. It was also small. It was also low.

The flat ground was the ground that could be built on. The hillsides were steep. The slopes rose at grades that made construction expensive and difficult. The valley floor was the only practical ground for a town, for a mill, for a rail yard. So the town built on the valley floor. The mill built on the valley floor. The rail yard built on the valley floor. Everything that mattered was on the lowest ground.

The Cambria Iron Company owned the largest of the mills. It sat on the valley floor just above the confluence, where the two rivers came together. The company had chosen the ground deliberately. The furnaces needed water for steam. The forges needed water for cooling. The rolling mills needed water for the boiler feed. The rail lines needed the river grade to enter the valley. The river was a resource. The river was also a grade. The Pennsylvania Railroad followed the Little Conemaugh from the east. The Baltimore and Ohio followed the Stonycreek from the south. Both lines converged at Johnstown because the rivers had cut paths through the mountains. The rivers made the valley. The valley made the town. The town made the mill.

Cambria Iron was not a small enterprise. By 1888, it operated six blast furnaces, a Bessemer steel plant, three rolling mills, a rail mill, and a wire mill. It employed approximately 7, 000 men. The payroll circulated through Johnstown in a weekly stream. The men rented houses from the company or from private landlords who depended on the mill workers’ wages. They bought food from grocers who stocked against the shift schedule. They bought clothing from merchants who priced against the wage rate. The company store, the company housing, the company hospital — the institutions of the town were either owned by Cambria Iron or shaped by its presence.

The town grew to serve the mill. The population reached 30, 000 by 1888. The wards spread along the river flats. Cambria City, a working-class district, sat on the west bank of the Stonycreek, directly across from the mill. Conemaugh borough sat upstream along the Little Conemaugh. Millville sat between the mill and the hillside. Woodvale sat in the narrow ravine of the Stonycreek to the south, pressed between the stream and the hill. Prospect Hill rose above the mill. The higher ground held the houses of the managers and the owners. The lower ground held the houses of the workers. The lowest ground held the river.

The arrangement was not accidental. The furnaces needed the river. The boilers needed the river. The rail lines needed the river grade. The workers needed to live near the mill. The mill needed to be near the rail. The rail needed to be near the river. The logic was circular. The logic was also iron. Each link in the chain depended on the one before it. Each link fixed the town more firmly on the valley floor. Each link made the river more central and the hillside more marginal. The town could not move uphill without breaking the chain. The mill could not move uphill without losing the water. The rail could not move uphill without losing the grade. So nothing moved. Nothing moved up.

The valley received large amounts of runoff from rain and snowmelt. The mountains on both sides drained into the two rivers. The drainage area above Johnstown covered hundreds of square miles of steep, deforested slopes. The timber cut for the mines and the mills had stripped the hillsides. The stripped ground shed water faster than forested ground. The water came down the slopes. The water came down the gorges. The water came into the rivers. The rivers came into Johnstown.

The freshets were a known feature of the valley. The Conemaugh had flooded in 1808, in 1811, in 1816. The Little Conemaugh had flooded in 1832, in 1847, in 1865. The Stonycreek had flooded in 1867, in 1873, in 1875. The records existed. The records were in the newspaper files. The records were in the county histories. The records were in the memory of the older residents. The water came. The water went. The town dried itself and continued.

The habit of recovery was the hinge. Each freshet confirmed that the river was a nuisance. Each recovery confirmed that the nuisance was temporary. The town pumped the cellars. The mill swept the floors. The railroad cleared the debris from the bridges. The merchants replaced the stock. The cost was real. The cost was also bearable. The cost was absorbed as overhead — the price of operating in a valley that had water in it.

But the cost was bearable only because the water had a ceiling. The freshets of the 1880s reached a certain stage. The water rose to the boiler pads. The water rose to the cellar windows. The water rose to the plank flooring. It did not rise higher. The town’s estimate of the river’s capacity was based on what the river had done. The estimate did not include what the river could do. The estimate did not include the reservoir.

Fourteen miles upstream, the South Fork Dam sat across the Little Conemaugh. The dam held back a lake. The lake covered 450 acres when full. The dam stood 72 feet high. The dam stretched 931 feet from hillside to hillside. The dam held back a body of water that, if released, would send a wall down the fourteen miles of gorge between the dam and the city. The wall would not be a freshet. The wall would not rise to the cellar windows. The wall would not stop at the boiler pads.

The town did not calculate this. The town calculated the river as it was. The river was a stream that rose and fell. The dam was a structure on a map. The reservoir was a lake that the club used for sailing. The connection between the lake and the river was not made. The connection between the dam’s condition and the town’s safety was not made. The town looked at the river. The town did not look upstream.

The institutional structure did not require it to. No state agency monitored the dam. The Commonwealth of Pennsylvania had sold the reservoir and the dam to the Pennsylvania Railroad in 1857. The Railroad had sold it to private interests. The private interests had sold it to the South Fork Fishing and Hunting Club in 1879. Each transfer moved the structure further from public oversight. Each transfer treated the dam as real estate. The dam was not real estate. The dam was a mass of earth holding back a mass of water. But the law classified it as property. Property was bought and sold. Property was not inspected. Property was not regulated. Property was private.

The club’s members lived in Pittsburgh. They were steel men, coal men, railroad men, bankers. Their ledger recorded the costs of the clubhouse, the boats, the fish, the caretaker’s wages. Their ledger did not record the cost of the dam’s condition. Their ledger did not record the risk below the dam. The ledger was a private book of expenditures and pleasures, maintained by owners who lived far from the consequences their property might unleash. The ledger balanced because the risk was not in it.

The valley below kept its own ledger. The ledger of Cambria Iron recorded tons of rail, tons of steel, tons of pig iron. It recorded the wages of 7, 000 men. It recorded the output of six blast furnaces. It recorded the cost of coal and limestone and ore. It did not record the cost of the dam above. It did not record the cost of the water that the dam held back. The valley’s ledger was a ledger of production. The club’s ledger was a ledger of leisure. The two ledgers did not connect. The dam stood between them. The dam connected them physically. No ledger connected them financially.

The gap between the two ledgers was the gap that made the disaster possible. The club owned the dam. The club did not own the valley. The valley did not own the dam. The valley did not monitor the dam. The club did not monitor the valley. The two parties were connected by a river and separated by everything else. The river flowed from the dam to the valley. The risk flowed with it. The risk did not appear in either ledger. The risk was unaccounted for. The risk was a deferred-maintenance debt that no one had booked. The debt accumulated. The debt would come due.

The town’s experience with water taught it the wrong lesson. Each freshet was a data point. Each data point said the water came, the water rose, the water fell, the work resumed. The data points formed a pattern. The pattern said the river was manageable. The pattern said the pumps were sufficient. The pattern said the cellars would dry.

The pattern was built on the freshets of the 1880s. The freshets of the 1880s were river floods. They were not dam failures. The difference was the difference between a stream rising and a lake emptying.

The town’s data did not include a lake emptying. The town’s data could not include a lake emptying. No lake had emptied in the valley’s recorded history.

The dam had been there since 1853. The dam had held. The fact that the dam had held was taken as evidence that the dam would hold.

The fact that the dam had been altered was not known. The fact that the dam had been lowered was not known. The fact that the spillway had been narrowed was not known. The fact that the discharge pipes had been removed was not known.

The town did not know what it did not know. The town did not know because no one told it. No one told it because no one was responsible for telling it.

The institutional vacuum was complete. The Commonwealth of Pennsylvania had divested itself of the dam in 1857. The Pennsylvania Railroad had divested itself of the dam in 1875. The private intermediaries who held it between 1875 and 1879 had divested themselves of it by selling it to the club. The club had acquired it. The club had modified it. The club had not maintained it. The club had not inspected it. The club had not reported on it. No agency had required the club to do any of these things. No law had required the club to do any of these things. The dam was private property. Private property was unregulated. The dam held back a public hazard. The hazard was not regulated either. The hazard sat in the gap between the private ledger and the public risk. The gap was not an oversight. The gap was the structure.

The valley’s geography concentrated the risk. The narrow gorge between the dam and Johnstown compressed the channel. The river ran through a series of bends. The bends would slow a normal flood. The bends would not slow a wall of water released from a lake. The wall would fill the gorge. The wall would maintain its height. The wall would arrive at the valley floor with most of its energy intact. The valley floor was flat. The valley floor was low. The valley floor was where the town was. The valley floor was where the mill was. The valley floor was where 30, 000 people lived. The valley floor was where the risk would land.

The risk was not theoretical. The dam had been modified. The modifications had reduced its capacity. The modifications had reduced its margin. The modifications had been made by men who did not live below the dam. The modifications had been made for convenience. The crest was lowered to widen the carriage road. The spillway was narrowed. The discharge pipes were removed. Each modification reduced the dam’s ability to hold water. Each modification increased the probability of failure. The probability was not calculated. The probability was not discussed. The probability was not in the ledger.

The town below the dam was a town that had organized itself around iron. The iron required water. The water required the river. The river required the valley. The valley required the flat ground.

The flat ground was the low ground. The low ground was the dangerous ground. The dangerous ground was where the value was concentrated.

The value was enormous. The mill alone represented millions of dollars in furnaces, rolling stands, rail mills, wire mills, boiler houses, engine houses, rail yards, and finished stock. The town represented thousands of houses, shops, churches, schools, and the infrastructure that connected them. The value sat on the valley floor.

The risk sat on the hillside above. The two did not meet in any ledger. The two did not meet in any law. The two did not meet in any institution. The two met only in the river.

The river ran from the dam to the valley. The river connected what the institutions did not.

The freshets of the 1880s were the last warnings the town would receive. The warnings came from the river. The warnings said the water rose. The warnings said the water fell. The warnings said the town survived.

The warnings were read as reassurance. The warnings were not reassurance. The warnings were evidence that the valley was vulnerable.

The valley was vulnerable because it was low. The valley was low because the mill needed water. The mill needed water because iron required it. Iron required it because the economy required iron. The economy required iron because the nation was building railroads. The nation was building railroads because the nation was expanding.

The chain of causation ran from the expansion of the United States to the location of Johnstown to the vulnerability of the valley floor. The chain ran through the dam. The dam was the link that the chain did not account for.

The dam was the link that no one owned in any meaningful sense. The club owned the deed. The club did not own the risk. The risk owned the valley.

The men who pumped the cellars in March 1887 did not think about the dam. The foreman who logged the shutdown did not think about the dam. The merchant who swept the mud from his storefront did not think about the dam. The engineer who inspected the railroad bridge did not think about the dam. The burgess who walked the streets after the water receded did not think about the dam. They thought about the river. The river was what they could see. The river was what they could measure. The river was what they could pump. The dam was fourteen miles away. The dam was behind a hill. The dam was on a map. The dam was not in their lives. The dam was not in their ledger. The dam was not in their calculations.

The calculations were wrong. The calculations were wrong because they were based on the river. The river was not the only source of water. The dam was above the river. The dam held a lake. The lake was above the town. The lake was held by an earth embankment that had been lowered, patched, and narrowed.

The embankment was owned by men who did not live below it. The embankment was not monitored by any agency. The embankment was not inspected by any engineer. The embankment was not in any public record of risk. The embankment was in the deed book of Cambria County. The deed book recorded the transfer of property. The deed book did not record the condition of the property. The deed book did not record the capacity of the property. The deed book did not record the hazard of the property. The deed book was a ledger. The ledger was incomplete.

The town’s vulnerability was a function of its success. The mill had grown because the valley was profitable. The valley was profitable because the rivers provided water and grade. The rivers provided water and grade because the mountains had carved them. The mountains had carved them because geology had placed them there.

The chain ran from geology to geography to industry to vulnerability. The chain was not visible from the valley floor. The chain was not visible from the hillside. The chain was visible only from the dam.

From the dam, one could see the lake behind and the gorge below. From the dam, one could see the connection.

The club members who stood on the dam and looked at the lake did not look down the gorge. The townsmen who stood in the valley and looked at the river did not look up the gorge.

The gorge connected them. The gorge was empty. The gorge was the space where the risk would travel.

The risk traveled in May. The risk traveled in 1889. The risk traveled because the rain came and the dam could not hold it.

The dam could not hold it because the dam had been altered. The dam had been altered because the club did not need the margin. The club did not need the margin because the margin was not in the ledger.

The ledger was in Pittsburgh. The cost was in the valley. The cost was the town. The cost was the mill. The cost was the cellars that had been pumped and the floors that had been swept and the bridges that had been cleared and the lives that had been organized around the river. The cost was everything that the valley had built on the low ground. The cost was everything that the valley had built because the low ground was where the value was.

The cost was the deferred-maintenance debt. The debt had been accumulating since 1881. The debt was eighteen years of lowered crest and narrowed spillway and removed discharge pipes. The debt was booked nowhere. The debt would be paid in water.

The concentrated risk in the valley, with no authority watching the dam above, created a pressure that demanded attention. The town had its pumps. The mill had its boilers. The railroad had its bridges. The river had its freshets. The dam had its cracks. The dam had its leaks. The dam had its lowered crest. The dam had its narrowed spillway. The dam had its history of modifications that no engineer had approved. The dam had its history of warnings that no authority had enforced. The warnings existed. The warnings came from men who had seen the dam. The warnings came from men who had measured the dam. The warnings came from men who had calculated the dam’s capacity. The warnings were not in the ledger. The warnings were not in the town. The warnings were in the gorge, between the dam and the valley, traveling downstream.