The Inquest and the Engineers' Committee

A survivor told the coroner’s jury that water had been trickling through the dam’s face before noon on May 31. A club representative told the same jury that the rainfall was unprecedented and no dam could have withstood it. The two statements sat in the same transcript. They described the same structure on the same morning. They did not describe the same event.

The relief camps had barely closed before the flood’s center of gravity moved from charity to blame. The autumn of 1889 produced two instruments through which Americans would decide who, if anyone, was answerable for Johnstown: the coroner’s inquest in Cambria County and the investigating committee of the American Society of Civil Engineers. Together they converted a catastrophe into a case.

The inquest convened in Ebensburg, the county seat, in the weeks following the flood. Coroner Jeremiah McMullen impaneled a jury of six men. He gave them a charge: to determine the cause of death for those killed on May 31, and to ascertain whether that cause involved criminal negligence. The instrument was blunt. A coroner’s inquest could not indict. It could not fine. It could not imprison. What it could do was compel testimony under oath and publish a finding. That finding would carry no legal weight in a civil suit. It would carry enormous weight in the press and the public mind.

The inquest summoned three categories of witness. The first comprised men who had operated the South Fork dam: the club’s superintendent, its employees, the men who had maintained the structure. The second comprised engineers who had inspected it: representatives of the Cambria Iron Company, railroad agents who had watched the reservoir. The third comprised survivors who had seen the dam fail. Their accounts would be the first time most of this testimony entered a public record.

The club’s position emerged as a contest of explanations, not a confession.

The club’s position, stated and restated across multiple depositions, held that the rainfall on May 30 and 31 exceeded any reasonable expectation. No structure, built to any standard, could have retained that volume of water. The storm was an act of God. The dam’s failure was consequent. The argument required that the dam have been sound before the storm began. It required that no prior condition, no structural deficiency, had contributed to the collapse. It required that the alterations to the dam — the lowered crest, the altered spillway, the patched breach — either had not occurred or had not mattered.

The engineering record suggested otherwise.

Survivors described water moving through the dam’s structure hours before the final collapse. The water was not overtopping the crest. It was passing through the embankment itself, emerging from the downstream face. Such seepage indicated that the dam’s internal structure had already begun to fail. The water was not simply rising. The earth was giving way.

The club’s representatives could not account for this observation within their framework. If the dam was sound, and the storm was unprecedented, then the failure should have begun with overtopping. Water should have appeared at the crest first, then flowed over the top, eroding the downstream face. Instead, witnesses described the opposite sequence. Water appeared through the face before water appeared over the top. The structure failed from within before the structure failed from without.

The inquest pressed on this point. It pressed on the dam’s history.

Cambria Iron’s engineers had inspected the South Fork dam in the years before the flood. Their correspondence, now entering the record, documented concerns about the crest’s height and the spillway’s capacity. They had communicated these concerns to the club. The club had responded with assurances. The correspondence was specific. The concerns were technical. The spillway, as modified, could not pass the volume of water that a serious storm would deliver to the reservoir. The discharge pipes at the dam’s base, which had once allowed controlled release of water, had been removed or rendered inoperable. The crest had been lowered by as much as three feet to widen the carriage road. Each of these changes reduced the dam’s capacity to handle high water.

The club’s superintendent addressed these points. He testified about repairs. He described work done to the dam and the spillway in the years before the flood. He described the screens placed across the spillway to prevent fish from escaping the lake. He described the practice of hauling earth and stones to patch areas where water leaked through. He described these as maintenance. The inquest heard them as something else.

The testimony established a pattern. When water appeared through the dam’s face, the response was to pack earth against the leak. When the spillway clogged with debris, the response was to clear the screens. When the water rose, the response was to add material to the top. Each response addressed an immediate symptom. None addressed the underlying condition. The dam was not being maintained. It was being patched. The distinction mattered. Maintenance preserves a structure’s designed capacity. Patching conceals its deterioration.

The absentee ledger — the private accounting of enjoyment and gain kept by owners who never came near the danger their property bred — could not be opened by the inquest directly. The club’s members, the Pittsburgh industrialists and financiers whose names appeared on the deed, were not called to testify. The inquest could reach the superintendent. It could reach the employees. It could not reach the men who had bought the dam, altered it, and left its operation to hired hands. The legal structure of the club protected them. The corporation stood between the owners and the instrument. The inquest could document what the employees did. It could not document what the owners knew.

The railroad agents provided a different kind of evidence. The Pennsylvania Railroad’s telegraph line ran through the Conemaugh valley. Its operators had tracked the rainfall. Its agents had observed the reservoir’s level in the days before the flood. Their testimony established the rate at which the water had risen. That rate, combined with the known capacity of the spillway and the discharge pipes, allowed a rough calculation. The reservoir was filling faster than the dam could empty. The arithmetic was simple. The dam’s discharge capacity, as modified, was insufficient for the inflow. The question was hydraulics, not opinion.

The inquest could not perform that calculation with precision. It had no hydraulic engineer on the panel. It had a county coroner and six laymen. What it had was testimony — specific, concrete, and conflicting. The survivors described what they saw. The employees described what they did. The engineers described what they had measured. The club’s representatives described what they believed. The transcript grew. The contradictions accumulated.

The central conflict was temporal. The club’s case required that the dam fail during the storm’s peak. The witnesses suggested it had begun failing before. If the seepage through the face began hours before the final collapse, then the dam was already compromised when the rainfall reached its maximum. The storm did not break a sound dam. It finished breaking a damaged one.

This distinction mattered legally. An act of God breaks sound structures. A damaged structure that fails under stress raises a different question. Who damaged it? Who knew? Who was responsible for repairing it? The inquest could not answer these questions. It could establish that they were legitimate questions. That was its function.

The inquest’s finding, when it came, was cautious. The jury found that the deaths had been caused by the breaking of the South Fork dam. It found that the dam had been improperly constructed and insufficiently maintained. It did not name the individuals responsible. It did not recommend criminal charges. It forwarded its findings to the district attorney. The district attorney would decline to prosecute. The legal standard for criminal negligence was high. The inquest had documented deficiencies. It had not proven intent or recklessness in the criminal sense. The finding was a public record. It was not a legal consequence.

The inquest had done its work. It had converted private knowledge into public testimony. It had placed the dam’s history — the lowered crest, the altered spillway, the patched breach — into a sworn record. It had established the conflict between the club’s narrative and the physical evidence. It had not resolved that conflict. Resolution required a different instrument.

The American Society of Civil Engineers provided that instrument. On June 5, 1889, five days after the flood, the ASCE appointed a committee of four prominent engineers to investigate the cause of the disaster. The committee was led by James B. Francis, a hydraulic engineer best known for his work on canals, flood control, and dam construction, and a former president of the society. The other members were William Worthen, Alphonse Fteley, and Max Becker. The committee represented the profession’s internal response. Its members belonged to the same world as the men who had designed, modified, and inspected the South Fork Dam. They spoke the same technical language. They used the same methods. Their authority derived from professional competence, not legal standing.

The ASCE committee’s investigation was methodical. The members traveled to the dam site. They examined the remnants. They commissioned a topographic survey of the break and the surrounding terrain. They reviewed the original design from the 1840s and the modifications made after the club’s purchase. They interviewed eyewitnesses. They collected rainfall data. They performed hydraulic calculations. Their work was a technical autopsy. The question was not who was to blame. The question was what had happened, and why.

The committee’s composition mattered. Francis was not a prosecutor. He was an engineer. His career had been built on hydraulic computation. He had developed the formula for measuring water flow over weirs that still bore his name. He approached the dam as a hydraulic structure. He asked hydraulic questions. What was the reservoir’s capacity? What was the spillway’s capacity? What was the inflow during the storm? At what point did the water level exceed the dam’s ability to contain it?

These questions had answers. The answers were calculable. The dam’s original design, as built by the Commonwealth in the 1840s and 1850s, had included a spillway and a system of discharge pipes at the base. The spillway was designed to pass excess water. The discharge pipes allowed controlled release. Together, they constituted the dam’s safety system. They were the mechanisms by which the reservoir could be managed during high water.

The club had altered both. The spillway had been modified. Its capacity had been reduced. The discharge pipes had been removed. The crest had been lowered. Each alteration was documented. Each had a measurable effect on the dam’s capacity to handle storm inflow.

The committee’s calculations would show that the modifications reduced the dam’s discharge capacity by roughly half. This was not a marginal reduction. It was a structural transformation. A dam built to handle a certain volume of water had been altered to handle half that volume. The alteration was not accidental. It was deliberate. The crest was lowered to widen the carriage road. The discharge pipes were removed because they had deteriorated. The spillway was screened to retain fish. Each decision served a purpose. None of the purposes was flood safety.

The committee’s investigation took months. It overlapped with the inquest but operated on different principles. The inquest heard testimony and weighed credibility. The committee gathered data and performed calculations. The inquest was bound by legal procedure. The committee was bound by engineering method. The inquest asked who. The committee asked how.

The two inquiries reached different audiences. The inquest’s transcript was a public document. It was quoted in newspapers. It was discussed in editorial columns. It was read by people who had lost family members and wanted to know who was responsible. The committee’s report was a professional document. It was published in the proceedings of the ASCE. It was read by engineers. It was read by lawyers preparing suits. It was read by legislators considering dam-safety regulation. Its authority was technical. Its audience was specialized. Its conclusions would carry weight that the inquest’s could not.

The committee’s report was completed in 1890. Its release was delayed. Becker, the outgoing president of the ASCE, held the report. Andrew Carnegie, who had become the new president of the society in January 1890, gave the report to Becker to decide when to release it to the public. The delay was unusual. The report was finished. Its conclusions were known within the committee. The delay raised questions about the society’s willingness to publish findings that might implicate prominent industrialists — some of whom were connected to the club and to the ASCE itself.

The report, when it appeared, was measured in its language and devastating in its particulars. It described the dam’s condition before the flood. It described the modifications. It described the spillway’s inadequacy. It described the removal of the discharge pipes. It described the lowering of the crest. It described the practice of patching leaks with earth and stones. It described a structure that had been systematically stripped of its safety margins.

The report did not use the word negligence. It did not need to. The engineering facts constituted the argument. A dam designed with a certain capacity had been altered to a lower capacity. A dam designed with multiple discharge mechanisms had been stripped of them. A dam designed to withstand high water had been made unable to withstand high water. The storm was the trigger. The modifications were the cause.

The report addressed the rainfall. It acknowledged that the storm of May 30 and 31 was severe. It did not concede that the storm was unprecedented. It placed the rainfall within a range of known meteorological events. It argued that a dam built to its original specifications, with its original spillway and discharge pipes, could have withstood the storm. The modifications had made the difference. A sound dam would have held. The altered dam did not.

This finding directly contradicted the club’s position. The club had argued that no dam could have withstood the storm. The committee found that the original dam could have. The alterations, not the storm, had created the conditions for failure.

The contradiction was not merely academic. It shaped the legal landscape. If the storm was unprecedented and the dam was sound, then the flood was an act of God. No one was liable. If the dam had been altered in ways that reduced its capacity, then the alterations were the proximate cause. The owners who ordered the alterations were responsible. The distinction between these two positions would organize every lawsuit filed in the flood’s aftermath.

The inquest and the committee had produced complementary records. The inquest provided the human testimony. It documented what the employees did, what the engineers observed, and what the survivors saw. The committee provided the technical analysis. It documented what the modifications were, what they cost in capacity, and what they meant for the dam’s ability to withstand the storm. Together, they constituted the most comprehensive record of the disaster’s causes that existed.

Neither instrument produced a verdict. The inquest forwarded its finding to a district attorney who declined to act. The committee published its report in professional proceedings that had no legal standing. The documented fault — the lowered crest, the removed pipes, the inadequate spillway, the practice of patching rather than repairing — was now a matter of public record. It was not a matter of legal consequence.

The gap between documentation and consequence was the chapter’s central fact. The inquest and the committee had done what they could do. They had established what happened. They had established why it happened. They had established who had the authority to prevent it. They had not established a mechanism for holding anyone accountable.

The club’s members remained behind the corporate form. The inquest could not reach them. The committee could not name them. The legal system would have to do what the inquest and the committee could not. It would have to determine whether the documented negligence constituted liability.

The deferred-maintenance debt — the accumulated cost of neglected repairs that was now passing from the dam’s private owners to the public below — had been audited. The inquest had counted the cost in human testimony. The committee had counted it in hydraulic calculations. The debt was enormous. It was measured in more than two thousand dead, in a valley destroyed, in a city that had to be rebuilt from its foundations. The audit was complete. The collection was not.

The lawsuits began almost immediately. Survivors filed claims against the club. The claims cited the inquest’s findings. They cited the engineering record. They argued that the club’s negligence — the modifications, the neglect, the failure to repair — had caused the flood. The legal question was not whether the negligence was real. The inquest and the committee had established that. The legal question was whether the negligence could be attached to specific defendants. Could the corporate form protect the club’s members? Could the legal standard for negligence reach men who had hired a superintendent and left the operation to him? Could the law connect the lowered crest to the dead in the valley?

These questions would occupy the courts for years. They would produce no successful verdict for the plaintiffs. The documented negligence would not become legal liability. The audit would not produce collection. The debt would remain with the valley.

The inquest transcript sat in the county courthouse in Ebensburg. The ASCE report sat in the society’s proceedings. Both were public. Both were available to anyone who could read them. Both documented the same facts. The dam had been altered. The alterations had reduced its capacity. The reduced capacity had caused the failure. The failure had killed more than two thousand people.

The documents were complete. The consequence was not. The fault was documented. The liability was not attached. The gap between the two would define everything that followed.

The coroner’s finding and the engineers’ report traveled together through the autumn of 1889 and into the winter. They moved from the hearing room and the survey site to the law offices where attorneys prepared their complaints. The testimony of the survivor who saw water trickling through the dam’s face and the calculation of the engineer who measured the spillway’s capacity entered the same briefs. They sat side by side in the legal record. They described the same failure. They named no one who could be made to pay for it.