by Jon Wilkman
The fourteen years between the first and last edition of Man-Made Disaster allowed Charley time to write more books about the West and make a few technical corrections and refinements in his first one. It also provided an unexpected opportunity to expand the human aspect of the story. In 1963, the author received an unexpected letter from a woman who identified herself as Tony Harnischfeger’s daughter. Outland knew of the dam keeper’s young son, Coder, but despite his years of research, he had never heard of a daughter.
Gladys Antoinette (Toni) Harnischfeger was born near Jawbone Canyon, where her father worked as a watchman, on alert for dynamite attacks. A close look at a photograph of the family at Jawbone Canyon shows her standing beneath her father and mother (see page 86). She stayed with her mother when her parents separated and her father moved to San Francisquito Canyon. When she showed up at Outland’s apartment in Santa Paula, she was wearing a bright red dress and appeared to be about forty. The surprised historian, who had an eye for the ladies, thought she was beautiful.
Toni filled in the blanks in her father and mother’s life before the collapse of the St. Francis Dam. It was not a happy story. In 1928, a reporter for the Los Angeles Record couldn’t resist a touch of melodrama when he described Gladys Harnischfeger’s reaction to the death of her estranged husband and young son. “Can’t I have the body of my little boy?” she was quoted as pleading. “Can’t I bury him with tender hands and place my flowers on his little coffin?”24
The bodies of Tony and Coder Harnischfeger were never found. Gladys received $11,000 to hold in trust for her daughter, and the two dropped out of sight. When Toni was eighteen, legally able to claim her father’s death benefit, she learned that her mother had spent it. It wasn’t a surprise. Shortly after she received the settlement, Gladys moved to Los Angeles, began to drink heavily, and supported herself by accommodating amorous men in seedy downtown hotels. Occasionally she attempted to rescue her sad life by becoming a born-again street preacher, but always reverted to old ways.
Toni told Outland that as soon as she was old enough, she left and went on her own. She eventually met her husband, Wayne Graham, and started a family. The couple owned a small grocery store in the Los Angeles port town of Wilmington. Toni had always been curious about her father and once tried to find the site of the St. Francis Dam, but the location had been well concealed for years.
On December 14, 1963, hundreds of thousands of Angelenos were tuned to live television. They watched as a V-shaped section of the foundation on the northeast corner of the Baldwin Hills Reservoir in Los Angeles eroded away and undermined an earthen embankment. A flood burst free and rushed down La Cienega Boulevard, only a few miles from Beverly Hills. With timely warnings to the large downstream population, only five would die.
In Wilmington, Toni Graham was bagging groceries for her friend Betty Edwards. They paused to watch the coverage on a TV positioned on a shelf behind the cash register. Toni casually mentioned that her father, Tony Harnischfeger, was a victim of the St. Francis Dam disaster of 1928. Taken aback, Betty told her friend that she’d recently read a book by a man named Charles Outland. She said Outland recounted the history of the tragedy and had included Toni’s father. Compounding the coincidence, Betty revealed that Thornton Edwards, the highway patrolman known as “the Paul Revere of the St. Francis Flood,” was her husband’s father.
Toni wrote Outland a letter and the author responded. A few months later, in 1964, she decided to visit the historian in Santa Paula. There were so many questions. After the two talked for the first time, Toni returned and Outland gave her a tour of the dam site, including his best guess where her father and brother’s cottage had been. When they finished tramping through the underbrush of San Francisquito Canyon, Toni quietly thanked her guide and left. Outland never heard from her again. Later, he learned she had died from a serious case of the flu in 1969. Toni’s mother Gladys passed away two years later.25 In a way, both were victims of the St. Francis flood.
In Outland’s book, mystery surrounds the memory of Tony Harnischfeger, but there is no ambiguity in the author’s portrayal of the exploits of motorcycle patrolman Thornton Edwards. However, even though Charley prided himself in finding the facts and telling the truth, he left the Edwards story incomplete.
Charley knew it is never easy to determine when a historical narrative begins and ends. He could have argued that unexpected events eleven years after the St. Francis Dam failure fell beyond the scope of Man-Made Disaster. But there was more to tell. After Thornton Edwards’s heroics in the early-morning hours of March 13, 1928, the grateful citizens of Santa Paula made him police chief. During the years that followed, many remembered the lawman fondly, but not all—especially members of the Mexican-American community.
In 1939, after seventeen years with local law enforcement, accusations emerged claiming that Edwards had become too big for his britches. There were rumors that the Paul Revere of the St. Francis Flood sometimes expected handouts from local merchants and used excessive force against Santa Paula’s growing Mexican-American community. It even was alleged he collected kickbacks from a local house of ill repute.
The burly lawman denied all charges and demanded proof. None was publicly presented. A few managers of citrus packing plants, the employers of many Santa Paula Mexican-American laborers, signed a letter of support for Edwards, and the Santa Paula Chronicle complained about a rush to judgment. It’s possible the accusations were unfounded, or it could have been that city leaders felt further exposure of the facts would prove embarrassing to Santa Paula’s small-town reputation—or, more troubling, lead to lawsuits.
Whatever the truth, on June 5, 1939, 150 people crowded into the Santa Paula City Council hearing room. During a contentious meeting, it took only ten minutes for a unanimous vote to relieve Thornton Edwards of his badge and duties as chief of police.26 The fired lawman didn’t contest the decision. A short time later, he returned to his first career as a B-movie actor, playing small parts in shoot-’em-up Westerns. Ironically, with his distinctive pencil mustache and his ability to mimic Spanish dialects, he was often cast as a Mexican bad guy. Perhaps Edwards’s most notable role was as the California highway patrolman who ushers the Joad family through a line of striking farm workers in the 1940 John Ford film The Grapes of Wrath, based on John Steinbeck’s novel.
Man-Made Disaster doesn’t deal with the controversy over the firing of Thornton Edwards even though, as the author would be the first to admit, legends have neat story lines and history can be ambiguous and untidy. Just as the flood spread into isolated canyons on the way to the ocean, the legacy of the St. Francis Dam disaster was not confined to issues of engineering; it had human dimensions that could be just as unexpected and long lasting.
When Outland prepared the 1977 revised and expanded edition of Man-Made Disaster, he again consulted the DWP and received a less defensive response. The Department’s reputation certainly had nothing to gain by probing old wounds, and perhaps a lot to lose, but after nearly fifty years, protective attitudes about a contentious past showed cautious signs of change. Even so, when Charley Outland died on March 28, 1988, only days after the sixtieth anniversary of the disaster he could never forget, he had unanswered questions, and no new facts to answer them.
Ten years before, in 1978, the year marking the fiftieth anniversary of the disaster, with Outland’s help a persistent journalism student helped organize a reunion of survivors, and the Santa Clarita Valley Historical Society erected a monument near the dam site. The reunion came and went and the marker eventually disappeared—either washed away by a flood or stolen. Another San Francisquito Canyon mystery. For nearly another twenty years, the story of the St. Francis Dam remained mostly forgotten local history, left out of the greater narrative of American social and technological history. In many ways, it was an abandoned puzzle and no one seemed interested in finding the pieces and reattempting to fit them together to see if the old picture had something new to show.
An exception was civil geological engineer J. David Rogers. Although not a trained historian, Rogers spent much of the 1980s applying the latest forensic methods to deepen an understanding of the failure of the St. Francis Dam. In 1992, he published an article that attracted some attention in national engineering circles.27 Three years later, he wrote a 109-page monograph that summarized his exploration of the disaster, the most extensive technical study since the investigative reports of 1928 and 1929.28 The article reviewed old studies and expanded past conclusions using computers and modern analytical methods known as rock and soil mechanics.
Rock mechanics is the theoretical and experimental study of how rocks and other solid materials like concrete react to forces applied to them. The discipline dates from the late nineteenth century, when geologists began to analyze how mountains like the Alps were formed, and to quantify the safety hazards of tunneling. Soil mechanics, the study of the interaction between soil and the surrounding environment, especially water, began to emerge in the mid-1920s. The first university course on the subject was offered at Caltech in 1934.
Both rock mechanics and soil mechanics are practical tools that contribute to an understanding of the dynamic relationship between the St. Francis Dam and the structure’s surroundings—information that is essential to understanding the failure, especially as it involved the forces of hydraulic uplift.
By 1928, dam designers appreciated the importance of geology, but rock and soil mechanics were relatively new analytical tools, and the full-time involvement of geologists on dam projects was unheard of. The first dam-site geologist was hired for the Hoover Dam in 1931, an addition to the planning and construction team that could be credited to the lessons of the St. Francis Dam.
The late 1920s was a period of transition in dam design and construction. Also, given William Mulholland’s resistance to new techniques and his belief that the foundation of the dam was on solid bedrock, analysis of rock and soil mechanics wasn’t applied to the St. Francis Dam’s factor of safety. By the 1990s, major changes had taken place in dam engineering, many of them made possible by the power of computer technology. Computers allowed J. David Rogers to utilize far more sophisticated rock and soil analysis when he sought to understand the complexities of the St. Francis Dam failure.
The use of computers vastly enhances older methods of dam design and construction, quickly performing complex calculations to measure and evaluate interaction of materials at a dam site and create models to test and evaluate alternatives. Although computer models are only as good as the data gathered to construct them, they can serve as engineering time machines offering a means to visualize the impact of future disasters—including the extent of a flooding created by the collapse of a dam that has yet to be built—and even anticipate loss of life based on existing census data from the area. Computer models can also revisit the past to reexamine the possible sequence of events that led to failures such as the collapse of the St. Francis Dam.
When Rogers joined the civil engineering faculty at the University of California, Berkeley, he had access to state-of-the-art computer modeling programs and talented graduate students to assist his quest to unravel the truth about what caused the collapse of the St. Francis Dam. Like the first forensic engineers who attempted to understand the failure in 1928, Rogers focused on the geology of the west and east abutments and the largest clues found at the dam site—the Tombstone and fragments of the dam left nearby and downstream. It was important to understand which pieces fell first and why, and, given the great size and weight of some of them, how they ended up so far downstream.
To answer this question, Rogers highlighted the fact that after the east-abutment landslides, the St. Francis floodwaters were thick with particles of schist, creating an extremely turbid, or muddy, flow. Turbid flows reduce the effective weight of submerged objects. In the case of the blocks released into the St. Francis flood, this reduction could be as much as 62 percent. As a result, fragments that weighed thousands of tons could be lifted and tumbled downstream, in some cases as far as three quarters of a mile.
Focusing on the abutments, the science of soil mechanics was useful in understanding what had occurred in the red conglomerate found in the upper west abutment, especially along the earthquake fault. But following Bailey Willis’s lead, Rogers was most interested in the east side. Here he applied “keyblock analysis,” a computer-aided approach developed at Berkeley in the 1980s.
Rogers divided the geology of the east abutment into wedge-shaped segments, the same way designers segment the cross section of a dam to evaluate stresses and determine a safe height and width. He wanted to study how uplift reduced friction against the Pelona Schist, which Bailey Willis had showed was part of paleo-landslides at the site. Rogers’s attention was drawn to an area of the lower east abutment that 1928 investigators had initially been unable to account for: the mysterious “missing section.”
Applying the analytical power of keyblock analysis, Rogers discovered that this unaccounted-for area of the dam would have become unstable when the reservoir of the St. Francis Dam rose to within seven feet of the spillways. On March 12, the water in the reservoir was only three inches from full capacity. Rogers concluded that a failure in this vulnerable portion on the east side of the structure was the place where the St. Francis Dam first began to break.
An admitted admirer of William Mulholland, Rogers acknowledged that the Chief accepted responsibility for the failure, but he argued that the understanding of uplift was still evolving in the 1920s and the existence of paleolandslides at the site was missed, not only by Mulholland but every post-collapse investigator, with the exception of Stanford geologist Bailey Willis. Articles in the press suggested this lessened the guilt, or even exonerated the Chief. Mulholland’s granddaughter and biographer, Catherine, welcomed a potential pardon, but two traditional historians disdainfully disagreed29 and other civil engineers weren’t willing to be lenient.
In the end, I believe the truth is not an excuse and is hauntingly ironic. When Bill Mulholland first arrived in Southern California in 1877 he probably passed through San Francisquito Canyon on his way to Los Angeles and his life’s work. Nearly thirty years later, he spent months living on the canyon floor during construction of the Owens River Aqueduct, and during the 1920s he repeatedly visited as the St. Francis Dam rose ever higher. The old man prided himself on his knowledge of field geology and was aware that San Francisquito schist was susceptible to slides, even if they weren’t ancient ones. Despite this, as Mulholland admitted at the Coroner’s Inquest, “We overlooked something here.”30 Something was indeed overlooked, but it wasn’t impossible to know or see.
Rogers points out that active involvement of geologists in dam planning and construction was rare in 1928, but professional awareness of the importance of geological analysis and the application of new countermeasures against uplift were far from absent. Despite this, after more than a half century of great achievements, acclaim, and controversy, Bill Mulholland believed he had long ago learned enough, and there was no one more powerful to make him act otherwise. As a tragic result, in addition to inadequacies in the St. Francis Dam’s placement and design itemized by Rogers and others, flaws in the Chief’s commanding personality, like unseen concrete cracks, contributed to the failure, along with hydraulic uplift and sliding ancient hillsides.
Whatever the disagreements concerning J. David Rogers’s opinion about the extent of William Mulholland’s culpability, the results of Rogers’s decades of computer-assisted forensic investigations are invaluable. They offer the most complete and convincing step-by-step scenario to describe and understand the collapse of the St. Francis Dam.
Rogers believed the cracks that appeared in the months before the failure were caused by contraction, as Mulholland assumed—but, as C.E. Grunsky concluded, the fissures also indicated stresses resulting from movements in the east abutment that increasingly compromised the curved concrete wall. In support of this, Rogers noted that
drivers traveling on the road that had been graded into the schist above the reservoir noticed small slides. Only three and a half hours before the failure, one traveler was forced to ease his vehicle over a twelve-inch drop in the road about one hundred feet upstream. The east abutment was in the midst of a slow motion slide exacerbated by the dynamics of water and concrete that would end in total collapse, bringing down the St. Francis Dam.
The missing section bursts.
Following J. David Rogers’s reconstruction, after an ominous prelude of cracks and slumping soil, the tragedy of the St. Francis Dam began suddenly around 11:57:30 when the pressure on the “missing section” caused the concrete to burst open. A massive landslide on the east side of San Francisquito Canyon follows, severing a Southern California Edison power line, causing a blackout in BPL Powerhouse 1. Using the block numbers assigned by the Governor’s Commission (see page 196), Rogers continues the forensic timeline.
When the landslide dumps a massive amount of soil into the St. Francis Reservoir, suddenly rising water levels create a wave that surges toward the west abutment. Virtually simultaneously, as the east hillside continues to fall, the floodwaters become increasingly viscous, and the soil creates a temporary earthen dam, slowing the outflow. By now, the Stevens Gauge is recording a precipitous fall in the reservoir level as water escapes the broken confines of the St. Francis Dam.