Recorded by the Seismograph

The seismograph at the University of California’s Lick Observatory began its trace at 5:12: 05. The needle, recording on smoked paper wrapped around a rotating drum, scratched a sharp vertical displacement that marked the arrival of the primary wave. The instrument had been installed in 1888, one of the standard Ewing vertical-motion seismographs that the university maintained to track the Pacific Coast’s seismic activity.

At the Student Observatory in Berkeley, the standard clock stopped at 5:12: 03. Professor A.O. Leuschner would later note in the Berkeley Reporter that this clock provided the best record of the beginning of the heaviest shocks, its mechanism jarred to a halt by the initial jolt, while less severe shocks were recorded some 35 seconds earlier.

The difference between these two times—two seconds, perhaps three—would become a small fixed point in the chaos that followed, a measurable instant when the accumulated strain of the San Andreas Fault released its energy upward through thirty feet of alluvial soil and into the structures of a sleeping city.

Forty-two seconds. That was the duration of the strong shaking, though survivors would swear it lasted minutes. The foreshock had come twenty to twenty-five seconds earlier, a brief warning that woke some sleepers and allowed them to rise before the main shock struck. Those who experienced both reported the foreshock as distinct, a sharp, single jolt, followed by the sustained violence of the principal earthquake. The seismograph at Lick Observatory continued its trace through the entire sequence, producing a record that geologists would study for decades, a paper ribbon of evidence that contradicted the simpler story the city would later tell itself.

The rupture propagated northward and southward from the hypocenter, which modern estimates place near Daly City or slightly offshore. The fault displacement, measured later by the State Earthquake Investigation Commission at points where roads, fences, and stream channels had been sheared, averaged eight to ten feet of horizontal offset. At a location twenty miles northwest, the ground had shifted twenty feet. The energy released, calculated retrospectively, would register approximately 7.9 on the moment magnitude scale that seismologists had not yet invented. The city above knew none of this. What it knew was motion: sudden, violent, and sustained.

In the South of Market district, the damage began immediately. This neighborhood housed the city’s working population in wood-frame lodging houses and cheap hotels, structures built rapidly after the 1890s boom and constructed without the masonry that would have required more expensive foundations. The ground here consisted of fill material, sand, debris, and organic matter dumped into the former marshlands of Mission Bay. When the seismic waves reached this saturated soil, they amplified.

Buildings that had stood for years tilted, twisted, and collapsed. The Brunswick Hotel at Third and Mission, a four-story wood structure, fell outward into the street. The Valencia Street Hotel, a three-story lodging house, dropped at its southeast corner, crushing the ground-floor rooms while leaving the upper floors angled at forty degrees. Survivors would later describe the sensation of the floor becoming a slope, of sliding toward windows that had become floors, of the strange silence that followed the initial crash.

At approximately 5:12: 30, the shaking found the California Hotel on Bush Street. This was a substantial building, five stories of brick and stone with ornamental cornices, constructed in the 1880s when such solidity represented permanence. On its roof, the fire department maintained Engine House No. 2, quarters for Chief Engineer Dennis T. Sullivan and his wife, Margaret. The chief had spent the evening reviewing the department’s preparedness, satisfied that his men and equipment stood ready for whatever the city might require. He had retired around eleven o’clock. The earthquake woke him with the first jolt. He was rising from bed when the brick chimney of the adjacent California Hotel collapsed through the roof of the engine house.

The debris struck Sullivan in the head and chest. His wife, thrown to the floor by the shaking, found him buried under masonry and timber. She managed to clear his face and summon help from the street below. Firemen from neighboring companies arrived to extract him, a process that took nearly an hour given the instability of the damaged structure. They carried him to the Receiving Hospital on Gough Street, where surgeons determined that his skull was fractured and his chest crushed. He remained conscious intermittently, asking after his department, before lapsing into the coma from which he would not emerge. He would die at 1:30 p.m. On Saturday, approximately eight hours after the earthquake, without having issued a single command regarding the fires that would consume his city.

The loss of Sullivan in the first minutes meant that the fire department entered its greatest trial without unified command. The assistant chief, technically next in line, found himself attempting to coordinate from the department’s central station at 410 Bush Street while the telegraph and telephone systems failed around him. The central station’s own chimney had collapsed, damaging the roof and blocking the apparatus doors. The assistant chief would later report that he sent runners to assess damage and locate available companies, but the destruction of communication infrastructure meant that these messages traveled at the speed of men on foot through streets already filling with debris and dazed civilians.

The damage to City Hall registered simultaneously. This building represented the city’s civic pride, a massive Beaux-Arts structure occupying two full blocks between Larkin and Polk Streets, constructed over twenty years and still incomplete in 1906. The central dome, rising 307 feet above the ground, had been dedicated only six years earlier. The earthquake sheared the dome from its supports; it collapsed into the building’s interior, destroying the Board of Supervisors’ chamber and the mayor’s offices below. The structure’s steel frame, innovative for its time, twisted under the lateral forces. The stone facing, inadequately anchored to this frame, peeled away in sheets.

The Hall of Records, adjoining City Hall and containing the city’s vital documents—property deeds, birth and death certificates, court judgments—suffered partial collapse of its upper floors. Clerks who had spent years organizing these records would spend the next three days attempting to salvage them while fire approached from three directions.

The water system failed in the same minutes that took Sullivan and cracked City Hall. This was not a single catastrophic break but a pattern of ruptures at critical junctions, the consequence of seismic waves traveling through different soil types and the brittle cast-iron pipes that the Spring Valley Water Company had laid over decades. At Fourth and Howard Streets, a 30-inch main ruptured. At Twelfth and Howard, another 20-inch line split. At Hayes and Fillmore, the junction of the University Mound and Sunset Reservoir supply lines, multiple fractures occurred. The total number of breaks would later be estimated at between 300 and 350, though precise accounting was impossible given the subsequent fires and the chaos of repair efforts.

What mattered in the first minutes was not the statistical total but the functional reality: water pressure dropped citywide as the network depressurized through these breaches. Fire hydrants that had delivered 80 pounds of pressure at 5:11 a.m. Delivered nothing at 5:15. Chief Engineer William H. Schussler of the Spring Valley Water Company, roused from his home on Pacific Avenue, would later testify that he reached the company’s primary valve house at 5:45 a.m. To find that the main supply from the reservoirs had already been shut down automatically by the loss of pressure. The system had protected itself by isolating its sources, but in doing so it had abandoned the city.

The gas system failed in parallel. The San Francisco Gas and Electric Company maintained 520 miles of distribution mains, predominantly cast iron, with service lines of wrought iron or lead extending to individual buildings. These pipes fractured at joints and weak points, releasing unburned gas into the ruptured structures. The company’s central station at First and Harrison, which generated the coal gas that supplied the network, suffered damage to its retort house and holder structures. The superintendent, arriving shortly after 5:30 a.m., found that the main valves could not be fully closed due to debris blockages. Gas continued to flow into the broken distribution system, seeking ignition sources in a city where chimneys had collapsed, stoves had overturned, and electrical systems had shorted.

The electrical infrastructure added its own failures. The California Gas and Electric Corporation’s power lines, strung on wooden poles throughout the city, snapped or grounded against damaged structures. Transformers exploded. The city’s new electric streetcar system, which had begun converting from cable operation only three years earlier, lost power simultaneously with the loss of its physical tracks, which buckled and fractured in the South of Market and along the Embarcadero. The United Railroads power house at Hayes and Fillmore, which supplied current to the city’s electric lines, shut down automatically when its governors detected the frequency fluctuations caused by earthquake damage to its distribution network.

The telephone and telegraph systems, which might have permitted coordination despite these physical failures, collapsed under the combined damage to their lines and central exchanges. The Pacific Telephone and Telegraph Company’s main exchange at 145 New Montgomery Street, a modern steel-frame building, survived structurally but lost power and suffered damage to its switchboard equipment. Operators who reported for the morning shift found instruments dead or sparking. The telegraph lines that connected San Francisco to the outside world, Western Union and the Southern Pacific’s private system, suffered multiple breaks where their poles had fallen or their cables had been sheared by falling masonry. The first news of the disaster to reach the outside world would come via wireless telegraphy from a coastal station operated by the United States Navy, which picked up fragmentary reports from ships in the harbor and relayed them to Mare Island and then to Washington.

The human experience of these forty-two seconds varied with location, construction, and circumstance. John W. Nourse, a resident of the Western Addition, described being woken at quarter past five by shaking that made his house seem like it would break all to pieces.

He looked from his window to see the houses on the other side of the street rocking like ships in a storm, a perception shared by numerous survivors who reported the visual illusion of fluid motion in solid structures. In the densely packed tenements of Chinatown, where wooden buildings stood four and five stories tall with only narrow alleys between them, the earthquake produced immediate structural failures that trapped residents in collapsed upper floors. The Chinese Hospital on Sacramento Street, a three-story brick building completed only two years earlier, suffered severe cracking of its walls and displacement of its floors, forcing the evacuation of patients who had been housed there as an alternative to the crowded conditions of the neighborhood’s boarding houses.

The financial district, built on firmer ground and with more substantial construction, experienced damage that was severe but less immediately catastrophic. The Mills Building, a steel-frame skyscraper completed in 1892, stood with its frame intact but its stone facing extensively cracked and displaced. The Palace Hotel, the city’s premier hostelry, lost portions of its interior partitions and suffered damage to its elaborate glass dome, but remained structurally sound. The Merchants Exchange Building on California Street, with its massive trading floor and ticker systems, saw its stock quotation board collapse and its electrical systems fail, but the building itself required only evacuation, not abandonment.

The waterfront presented a distinct pattern of damage. The filled ground along the Embarcadero, material dumped to extend the city’s flat land into the irregular shoreline of Yerba Buena Cove, liquefied under the seismic shaking. Piles that had supported piers and warehouses tilted or sank. The Ferry Building, the city’s transportation hub with its 245-foot clock tower, survived with damage to its interior finishes and the stoppage of its clock at 5:12, but the structures along the adjacent wharves settled unevenly, cracking their wooden decking and tilting their transit sheds. The Southern Pacific’s ferry slips, from which commuters departed for Oakland every fifteen minutes during rush hours, suffered damage to their aprons and waiting rooms that would interrupt service for hours.

The military installations experienced the earthquake as both physical event and command challenge. At the Presidio, the Army’s coastal defense post at the city’s northern tip, the shaking damaged barracks and administrative buildings but left the gun emplacements and ammunition magazines intact. Brigadier General Frederick Funston, commanding the Department of the Pacific, was roused from his quarters at 5:13 a.m. By the motion and the sound of falling plaster. He would later report that he dressed and proceeded to his office, finding that telegraph communication with Washington had been severed and that his primary responsibility, coastal defense against naval attack, appeared unaffected by the seismic event. The Presidio’s water system, fed by its own reservoirs, maintained pressure. Its garrison of approximately 1, 600 officers and men stood ready for orders that had not yet been formulated.

At Fort Mason, the Army’s supply depot adjacent to the waterfront, the earthquake damaged warehouses and collapsed a portion of the coal storage facilities. The quartermaster, responsible for supplies to all Pacific Coast installations, faced immediate losses of stored provisions and the disruption of his distribution network. At the Benicia Arsenal, across the bay, the shock was felt but caused no significant damage; this facility would become crucial in subsequent days as the source of explosives for the firebreak operations.

The civilian institutions of order responded with the resources available to them. Mayor Eugene Schmitz, residing at 2816 Fillmore Street in the relatively undamaged Pacific Heights district, was woken by the earthquake and proceeded by carriage toward the city center, finding his path blocked by debris and crowds. He would later claim to have reached the Hall of Justice on Kearny Street by 6:00 a.m., though witnesses placed his arrival closer to 7:00.

The Board of Supervisors, whose regular meeting was scheduled for that morning, could not convene at City Hall; those members who attempted to reach the building found it evacuated and cordoned. The city’s judiciary, operating from the ruined Hall of Justice and the damaged Federal Building, faced the immediate problem of prisoners: the County Jail on Broadway had suffered structural damage that required evacuation of its inmates, who were marched under guard to the Broadway pier for temporary confinement on a moored barge.

The medical system activated in fragments. The Receiving Hospital, where Sullivan had been taken, found itself overwhelmed within the first hour by casualties from the collapsed buildings of the South of Market. The city’s other hospitals, St. Mary’s, French, Lane, and the German Hospital, suffered damage that reduced their capacity while increasing their demand. The medical school at Cooper Medical College, with its associated hospital on Sacramento Street, evacuated patients when its structure showed dangerous cracking. The Red Cross, which maintained no permanent San Francisco chapter, had no organized presence to coordinate the improvised aid that began flowing from private citizens and neighborhood organizations.

The transportation network, beyond its immediate physical damage, faced the functional problem of a population in motion. Residents of damaged buildings poured into the streets, many in nightclothes, carrying what possessions they could seize. The streetcar system, already crippled by power failures and track damage, could not absorb this movement. The cable car lines, which operated on grip mechanisms powered by central engines, maintained limited service on those lines where the tracks remained intact and the cables had not snapped, but the density of pedestrians and debris in the streets reduced their speed to a crawl. The ferry system, which would become the primary evacuation route in subsequent days, operated at reduced capacity due to damage to its San Francisco terminals and the disruption of its Oakland connections.

The fire department, in these first minutes, attempted to function without its chief and without water. Individual companies, responding automatically to alarms that were still being transmitted by the mechanical street-box system, reached fires that they could not extinguish. At 5:22 a.m., according to the department’s later chronology, the first alarm was transmitted from the vicinity of Hayes and Gough Streets, where a chimney collapse had ignited the roof of a private residence. Engine Company No. 25 responded with its steam pumper, laid hose to the nearest hydrant, and found no pressure. This pattern would repeat across the city: companies arriving at ignition points, connecting to dead hydrants, attempting to draft water from cisterns or the bay, finding their equipment inadequate to the scale of outbreak.

The cistern system, which Sullivan had championed as backup to the pressure mains, proved insufficient. The city maintained 150 cisterns with total capacity of approximately 10 million gallons, distributed through the downtown area. These had been filled in preparation for the summer fire season. But the cisterns were designed to supplement, not replace, the pressured system; their locations did not correspond to the points of ignition; and their access points were often blocked by debris or occupied by refugees. The first companies to attempt cistern drafting found that their hoses, designed for hydrant pressure, could not effectively lift water from below-grade reservoirs.

The gas-fed fires began in the same minutes as the structural ignitions. At 5:25 a.m., according to company records, the first confirmed gas explosion occurred at a commercial building on Mission Street near Third. The San Francisco Gas and Electric Company’s emergency crews, attempting to shut down the distribution network at its district valves, found that seismic damage had jammed or broken many of these controls. The gas continued to flow, seeking the ignition sources that the earthquake had scattered: overturned stoves, exposed electrical arcs, the thousands of chimneys that had collapsed into fireplaces still holding the previous evening’s embers.

The electrical fires added a third category of outbreak. The short-circuiting of damaged power lines produced arcs that ignited adjacent materials. The California Gas and Electric Corporation’s emergency crews, like their gas company counterparts, attempted to isolate their network but faced the same problems of access and equipment damage. The combination of these three ignition sources, structural ignitions from collapsed chimneys and stoves, gas-fed fires from broken distribution lines, and electrical fires from shorted circuits, produced a pattern of outbreak that was simultaneous, scattered, and beyond the capacity of any response system to address.

The seismograph at Lick Observatory continued to record aftershocks through the morning. These were numerous but diminishing; the major energy release had occurred in that initial forty-two seconds. The smoked paper trace showed the main shock’s complex wave train, the surface waves that had produced the sustained violent motion, the gradual attenuation as the rupture propagation ceased. The instrument had performed its function: it had recorded, with the precision of mechanical inscription, an event that human perception and memory would reconstruct with inevitable distortion.

The stopped clocks provided their own record. At the Ferry Building, the tower clock halted at 5:12. At the jewelry store of J.C. Ghirardelli on Market Street, a display clock stopped at the same minute. In private homes across the city, mantel clocks and pocket watches recorded times between 5:12 and 5:14, depending on their location relative to the rupture propagation and their own mechanical sensitivity. These material witnesses, like the seismograph trace, would later serve as correctives to the more dramatic accounts of survivors who experienced the earthquake as an extended, almost eternal duration.

The physical preconditions for the conflagration were established in these minutes. The water system had been breached at multiple points, its pressure lost, its sources isolated from its distribution network. The gas system continued to feed fuel into damaged structures. The electrical system produced ignition sources even as it failed to deliver power. The fire department had lost its chief and its capacity for coordinated response. The civilian authorities were disconnected from their administrative centers and from each other. The communication infrastructure that might have permitted organized reaction had been severed. The city stood with its protective systems disabled, its ignition sources multiplied, and its population in uncoordinated movement through streets that would soon become corridors of fire.

In the rubble of the California Hotel, where Sullivan lay crushed, the gas lines of the damaged structure had already begun to leak. The same was true at City Hall, at the Hall of Records, at the collapsed lodging houses of the South of Market. A ruptured water main or leaking gas line in the rubble handed off the physical preconditions, scattered ignition and no water, that created the opening for disorganized response.