Official After-action Reports from Army

Seen from above, there were two cities at once — the one that would emerge from the transformation, and the one that had existed at 5:12 a.m. on April 18 — and every decision about what stories to tell and what losses to bury would be measured against that comparison. Across the whole map of ruin, on the morning of April 21, the third day, that accounting began in earnest—not with ceremony but with the scratch of pen on paper in a tent pitched among the ruins, and with the last guttering flames where men still fought for buildings that no longer stood.

At the northern edge of the destruction, along the waterfront below Telegraph Hill, a detachment of soldiers from the 22nd Infantry worked methodically through the smoldering wharves. They had no orders now, no central direction. The telegraph lines were down, the runners had stopped coming, and the officers who might have commanded them were occupied with maps and requisitions three miles south. The men moved from pier to pier with shovels and burlap sacks, beating out sparks that reignited in the sea wind, dragging charred lumber into the water where it hissed and sank. They had been at this since midnight. Their rations had run out. One soldier, Private First Class Thomas Burns, would later record in a letter home that they had stopped expecting relief and started expecting simply to finish—to reach a point where there was nothing left that would burn.

That point came shortly after dawn, when the wind shifted and the last visible flame at Pier 33 collapsed into ash. The men stood in the sudden stillness, listening to the collapse of cooling timbers, and then began the walk south toward the Presidio without looking back.

Three miles away, in a requisitioned office on the grounds of the Presidio itself, a captain from the Department of California sat at a field desk compiling the first consolidated after-action report. The document would eventually run to forty-seven pages of single-spaced memoranda, but its essential work was already visible in the officer’s method: the translation of three days of chaos into categories—troops deployed, rounds expended, property destroyed, civilians assisted. Outside his tent, the morning fog was burning off, and through the open flap he could see the refugee columns moving along the roads toward Golden Gate Park. He did not record them.

His instructions, cut by General Funston himself, directed him to emphasize the successful cooperation of military and civil authorities in the suppression of fire and the preservation of order. The earthquake appeared twice in his draft, both times as a contributing factor to conflagration conditions. The word earthquake did not appear in his summary of damages. This omission aligned with the widespread practice of insurers to indemnify properties from fire but not from earthquake damage, a fact confirmed by a Boulder Creek resident writing five days after the shock that companies “will insure against fire, will not insure against earthquakes.”

The omission was not yet deception. It was the first stage of narrative triage: the selection of which facts could be managed and which must be left to the ruins themselves.

The fire, meanwhile, was completing its own conclusion without regard to any report. At the western barrier of Van Ness Avenue, where Major General Adolphus Greeley’s engineers had spent two days dynamiting mansions and soaking the rubble with wine from the flooded cellars of Merchant Street, the thermal mass had finally spent itself. The avenue itself had become a firebreak not through any triumph of engineering but through the simple exhaustion of fuel. The grand houses that had lined its eastern curb were now a ridge of blackened stone and twisted iron, and beyond them the fire had found nothing dense enough to sustain its advance. The western side, where volunteer bucket brigades had fought through the second night, remained largely intact—though intact meant only that the buildings still stood, their windows blown out, their interiors waterlogged and looted.

Greeley walked this line shortly after 8 a.m., accompanied by a photographer from the Army Corps of Engineers. The general’s report, filed that evening, noted that the conflagration had been successfully arrested by military operations, a formulation that assigned agency where there had been mostly desperation. The photograph, preserved in the National Archives, shows Greeley in profile against the smoking ruin, his hand raised as if directing attention to something beyond the frame. What lies beyond the frame is the Fillmore District, untouched, and beyond that the Western Addition, where the fire had never reached. The general’s gesture creates a geography of salvation that the image itself does not support.

By mid-morning, the last isolated blazes had been reduced to smoldering heaps that required attention only from those with nowhere else to be. At the corner of Greenwich and Powell, where the fire had jumped the break at Lombard Street and consumed a block of frame houses before dying in the open ground of a demolished lot, a squad of firefighters from the remnant of Engine Company No. 2 stood amid the wreckage of their own efforts. They had no water. The cisterns were dry, the mains shattered, and the bay too distant for any bucket chain to reach.

Their chief had died in the first hour of the disaster, crushed when the California Hotel collapsed onto the fire station on Bush Street. Dennis T. Sullivan, fifty-four years old, forty years in the department, had been sleeping in the quarters adjoining the hotel when brickwork from the cornice fell through the roof of the engine house. He lived for four days, conscious for most of them, and died on April 22 without ever learning the full extent of what his city had become.

In 1909, the city would launch a fireboat bearing his name, and when San Francisco built a new residence for its fire chiefs, it too would be named after Sullivan. These memorials would preserve the man while the circumstances of his death—earthquake, not fire—would be gradually elided from official memory.

The firefighters at Greenwich and Powell did not know this. They knew only that their department had ceased to exist as an organized force, and that their presence at this particular corner was a choice made in the absence of any command. One of them, a hoseman named William McMullen, would later testify before the Board of Supervisors that they had remained to prevent reignition, though there was nothing left to reignite. The truth was simpler: they had nowhere to go, and the work, however futile, maintained a structure of purpose against the collapse of everything else. When they finally abandoned their post at noon, they walked south through streets that were already being renamed in the official mind—not as disaster zones but as cleared areas suitable for temporary structures and relief distribution.

The relief itself was becoming visible in the same hours that the last flames died. At Golden Gate Park, where the Army had established the largest of the refugee camps, soldiers from the 9th Cavalry were erecting tents in the meadow east of the polo field. The work had begun on the afternoon of the second day, and by the morning of the third, the camp held approximately 20, 000 people—though held suggests an order that did not exist.

The refugees had arrived in waves: first the property owners from the western districts who had loaded wagons with furniture and silver and established themselves with a sense of temporary inconvenience; then the renters from the South of Market, carrying what they could bundle in blankets; finally the displaced from Chinatown and the Italian North Beach, who came with nothing and found the best ground already occupied. The Army’s report would note satisfactory progress in camp organization, but the photograph taken at 10:30 a.m. Shows a different reality—blankets laid on bare ground, cooking fires in improvised pits, a line of women waiting at a water wagon that had not yet arrived.

General Funston visited this camp at 11 a.m., accompanied by Mayor Eugene Schmitz. The two men made a circuit of the meadow, speaking with refugees who approached them, accepting reports from officers who did not know what to report.

Funston’s manner, according to the Oakland Tribune correspondent who accompanied him, was businesslike, almost curt. He asked about sanitation, about water, about the disposition of the 9th Cavalry’s horses. He did not ask about the dead, or about the circumstances by which these particular 20, 000 had arrived here while others had found shelter in private homes or had not found shelter at all.

Schmitz, by contrast, appeared visibly affected, repeatedly shaking his head at the sight of families huddled under improvised shelters. The mayor had issued the shoot-to-kill proclamation on the first day, authorizing soldiers and police to fire upon looters; he had sat with the Committee of Fifty through two nights of deliberation; he had approved the dynamiting of the downtown. His emotion at the refugee camp was not recorded as contradictory, and the Tribune story treated it as natural—the human response of a leader confronting the cost of necessary measures.

The photograph of this visit, preserved in the Bancroft Library, shows the two men standing apart from each other, each surrounded by his own staff, each facing a different section of the crowd. Funston looks toward the western edge of the meadow, where the park’s natural forest begins. Schmitz faces east, toward the city. Between them, in the middle distance, a child sits on a blanket, looking at neither man. The image captures a division that would widen in the months to come: the general’s war against disorder, the mayor’s struggle to reclaim political authority from the emergency he had helped authorize.

By early afternoon, the official narrative was consolidating faster than the ruins were cooling. At the headquarters established in the Mechanic’s Pavilion on Grove Street—one of the few large buildings still standing in the civic center—the captain’s preliminary figures were being copied into telegrams to Washington. The message, sent at 2:15 p.m. And preserved in the War Department files, reported that the fire was now under control, estimated property loss at 200 million dollars, noted that troops were maintaining perfect order, and stated that civil authorities were cooperating effectively.

The estimate was low by half, the cooperation had been imposed rather than invited, and the fire was not under control but simply finished. Yet the telegram served its purpose: it established a version of events that could be cited, repeated, and eventually believed.

The War Department’s reply, received at 4 p.m., congratulated Funston on the successful management of a great emergency and authorized the continuation of martial law for as long as required by conditions.

The conditions themselves were shifting from combustion to exposure. As the thermal mass of the fire dissipated, the city revealed itself as a landscape of sudden temperature: black stone that held heat through the night and released it slowly, cellars that steamed with the condensation of vaporized water, collapsed buildings whose interiors remained ovens while their exteriors crusted with cooling ash. Survivors moving through this terrain developed a new vocabulary of hazard—not falling walls, which were obvious, but the collapse of fire-weakened floors, the release of trapped steam, the reignition of gas pockets in the rubble. The official reports would eventually classify these as post-conflagration risks, but in the hours after the main fire died, they were simply the conditions of movement through a transformed city.

At the Ferry Building, which had survived as the eastern anchor of the downtown, a group of perhaps two hundred people had gathered by mid-afternoon, not to cross the bay—they had no fares, and the ferries were running only for official traffic—but to wait. They had been waiting for three days, moving as the fire moved, and they had developed a collective patience that required no explanation. When a relief wagon from Oakland finally arrived at 3 p.m., the distribution of bread and coffee was accomplished in silence, without the pushing and shouting that had marked similar efforts on the first day. The waiting had become its own structure, and the food, when it came, was accepted as a continuation of that structure rather than its relief.

The true aftermath was not the fire’s end but the persistence of need beyond any institutional capacity to name it. The Army’s reports, compiled through the afternoon and evening of April 21, could account for troops and equipment, for miles of street patrolled and buildings dynamited. They could not account for the woman who approached the captain’s tent at 5 p.m., carrying a child who had not eaten in two days, asking not for food but for someone to tell her where to go. The captain directed her to the Golden Gate Park camp, though he knew the camp was full and that no systematic registration existed to prevent her from being turned away. His report noted that civilian inquiries were handled satisfactorily. The child was not mentioned.

The wind that had driven the fire for three days died at sunset, or shifted so slightly that its absence was felt as a change in pressure rather than direction. The sky, which had been yellow-gray with smoke since Wednesday morning, cleared enough to show stars in the eastern quadrant. The first natural darkness the city had experienced since the earthquake revealed how completely the urban light had been extinguished. Where 200, 000 electric lamps had burned on Tuesday night, perhaps two hundred now functioned, all of them in the Presidio or the western districts beyond the fire line. The darkness was total in the burned zone, broken only by the glow of cooling ruins and the scattered fires of refugees.

In this darkness, the last official measurement of the disaster was attempted. Lieutenant Colonel George H. Torney, the Army’s chief medical officer for the Department of California, led a party through the South of Market district with lanterns and a tally sheet, attempting to count the dead. The effort was abandoned after three hours. The ruins were too hot to search thoroughly, the bodies too burned to identify, the risk of structural collapse too great for systematic work. Torney’s report, filed the following morning, estimated not less than 300 dead in the district surveyed, with the caveat that complete accounting was impossible under present conditions. The number was meaningless except as a placeholder, a figure that could be cited in official correspondence while the true count—whether 500 or 5, 000—remained buried in the ash.

The fire was out, but survival had become the new crisis before the last embers were cold. At the Presidio, General Funston convened his staff at 9 p.m. For a reading of the consolidated reports, and the discussion turned not to what had been saved but to what remained to be done: the feeding of a population that would reach 200, 000 in the refugee camps, the prevention of epidemic disease in a city without water mains or sewerage, the maintenance of order in a landscape where property rights had been suspended and the distinction between salvage and looting had never been clearly established. The meeting ended with a telegram to the War Department requesting additional supplies, additional troops, and—significantly—additional authority to requisition property and labor. The emergency, which had begun as a response to fire, was now a permanent condition of governance.

In the camps, the night brought cold rather than rest. The Army had issued 10, 000 blankets by evening, enough for one in five refugees. The others improvised with curtains, carpets, clothing layered until movement became impossible. At Golden Gate Park, the temperature dropped to 42 degrees, and the morning would find frost on the tents of those who had them. The Examiner, which had resumed publication from a plant in Oakland, would describe this as the first peaceful night since the disaster, a formulation that treated the absence of flames as peace and the presence of freezing, hungry, homeless thousands as merely the new normal. The newspaper’s own reporters, sleeping in the offices of the Oakland Enquirer, knew better, but the story they filed was the story that could be printed: the fire defeated, the city saved, the work of rebuilding about to begin.

The rebuilding, however, required first a reckoning with what had been destroyed—and with what had been done in the name of salvation. That reckoning would not come immediately. In the hours after the fire died, the priority of survival overwhelmed every other claim: the need for water, for food, for shelter against the cold that the fire’s end had made suddenly pressing.

The official reports, compiled in tents and requisitioned offices, would establish the framework within which that reckoning would eventually occur, framing the disaster as a fire rather than an earthquake, as a challenge met rather than a catastrophe compounded, as a story with a clear conclusion rather than an ongoing crisis of authority and need. The fire was out. The city that remained was measured against the city that had been, and the comparison would shape every decision that followed—not through explicit choice but through the simple fact of what had been recorded and what had been left to the darkness, what could be named and what could only be endured.