The Week of Four Thousand Dead

The Bureau of Health bulletin for Saturday, October 12, 1918, arrived at the newspaper offices in the early afternoon. A single typed sheet, the carbon still faintly warm from the duplicating machine. The clerk who received it at the Public Ledger noted the figure in the final column: 289 deaths from influenza and pneumonia for the twenty-four hours ending midnight Friday. He set it on the city editor’s desk without comment. The editor glanced at it, turned it over, and reached for the afternoon’s wire copy. The bulletin did not appear in that evening’s paper.

On the same day, the Inquirer carried a full page of war news, a photograph of President Wilson reviewing troops in France, and a brief notice that the Fourth Liberty Loan subscription total had reached $6.9 billion nationwide. The paper’s evening bulletin had, twenty-four hours after the parade ended, described the event as “the first premonition of – victory” and “a great day in Philadelphia.” Now, the death toll from the epidemic, which had crossed two hundred cases daily for the first time that week, went unmentioned. The paper did note that the Shubert Theater had canceled its performance of The Squab Farm owing to the epidemic, and that the Drexel Theater on Market Street had followed suit. Small items, buried among the classifieds.

The silence was itself a datum. The city’s newspapers had covered the Liberty Loan parade with seventeen column inches of triumphal prose. Now, as the harvest of that gathering came in, they retreated into a vocabulary of adjournment and postponement. Meetings of the Rotary Club, the Kiwanis, the Sons of the American Revolution—all noted as indefinitely postponed. The society pages, which in September had chronicled the return of naval officers and the charity balls of the Main Line, shrank to a quarter of their former size. The Evening Bulletin, which maintained the most complete record of civic life, ran on October 12 a single paragraph stating that the Board of Health had issued a warning against unnecessary visiting. The warning itself was not quoted.

The week when the numbers became undeniable and simultaneously invisible. The Bureau of Health continued its daily bulletins, transmitted by telephone to the papers and posted at police stations and post offices. The Ledger would resume publication of the figures on October 14, but even then they appeared without editorial comment, as weather reports might appear, as facts without interpretation. The city had not yet developed a language for what was happening to it.

—-

Sunday, October 13. The mortality bulletin recorded 289 deaths again, the plateau suggesting not improvement but the maximum capacity of the recording system. The city’s physicians were now seeing patients in numbers that exceeded any previous experience. Dr. William H. Rorer, a general practitioner with offices on South Broad Street, later recalled that he made sixty house calls on Sunday alone, moving by automobile between homes where entire families lay prostrate. His account, published in a medical journal the following year, noted the characteristic symptoms: the sudden onset, the cyanosis that turned patients dark blue, the hemorrhage from nose and lungs. He did not record his own fatigue.

The hospitals had stopped admitting non-influenza cases three days earlier. Pennsylvania Hospital, which had maintained separate wards for the epidemic since October 3, now turned away patients at the door. The University of Pennsylvania’s hospital on Spruce Street reported 150 influenza admissions on October 13, with twenty-three deaths in the preceding twenty-four hours. The nursing staff, depleted by illness, worked shifts of sixteen hours. The Inquirer carried a small notice that the hospital’s training school had suspended classes and assigned all students to ward duty.

The streetcars ran on their regular schedules, but the Philadelphia Rapid Transit Company reported ridership down by 40 percent. Those who rode sat apart, when they could. The Bulletin noted that the company had ordered its conductors to refuse passengers who coughed without covering their mouths, a regulation impossible to enforce and rarely mentioned again. The cars were not disinfected. The company’s principal concern, evident in its correspondence with the city, was maintaining sufficient crew to keep the lines open.

Monday, October 14. The bulletin climbed to 376 deaths. The Ledger published this figure on its front page, in a single column between a report on the Meuse-Argonne offensive and a notice that the Wanamaker store had obtained a special shipment of fall coats. The placement was significant. The war remained the organizing narrative of the page, the epidemic an interruption of commerce. The paper did not attempt to relate the 376 deaths to the 118 deaths recorded on September 29, the day after the parade. No editorial voice drew the connection. The numbers accumulated without accumulation of meaning.

At the Navy Yard, the medical officer of the day recorded 47 new admissions and 12 deaths. These figures, reported separately to the Bureau of Navigation in Washington, were not included in the city’s bulletins. The dual accounting system—military and civilian—had created two separate epidemics, each with its own mortality curve. The Navy’s peak had come earlier, in the last week of September, when the receiving ship Blue Jacket had reported 600 cases among its crew of 1, 500. By October 14, the naval authorities believed they were past the worst. They were correct for their own jurisdiction. The city’s epidemic lagged by ten days, the interval between exposure and death now working through the civilian population that had gathered on Broad Street.

The volunteer network, which had organized itself in the cellars and parish halls of the previous week, reached its limit on October 14. The Visiting Nurse Society reported that its staff of 125 nurses had been reduced by illness to 87, and that these were working an average of seventy hours weekly. The Emergency Aid, a women’s organization founded for war relief, had shifted its entire operation to influenza care. Its headquarters at 1420 Walnut Street received 800 calls for assistance on October 14. The volunteers could answer 300. The remainder went into a logbook that grew by the hour, names and addresses that would never be visited.

Tuesday, October 15. The bulletin: 428 deaths. The Inquirer placed this figure on page three, under a headline stating that the City Health Board had taken steps to check the epidemic. The steps were not specified. The paper’s lead story concerned the capture of the French city of Laon by American forces. The epidemic received less space than the Society for the Prevention of Cruelty to Animals’ annual report.

The silence of the factories became audible. The Baldwin Locomotive Works, which had employed 16, 000 men in September, reported absenteeism of 35 percent. The Midvale Steel and Ordnance Company, manufacturing shell casings for the Western Front, shut two of its four furnaces for lack of workers. The federal government’s priority system, which guaranteed railroad cars and raw materials to war production, had no mechanism for replacing sick men. The Ordnance Department in Washington telegraphed inquiries about delivery schedules. The companies telegraphed back figures of incapacitation that the government had no category to receive.

At the Curtis Publishing Company, where the Ladies’ Home Journal and Saturday Evening Post were printed, the bindery operated with a skeleton crew. The October 19 issue of the Post would carry Norman Rockwell’s illustration of a boy in a soap-box derby, an image of normalcy composed in a city where normalcy had ceased. The editors at Curtis made no reference to the epidemic in their pages. Their readers, scattered across the country, received no news of what was happening in the city where their magazines were made.

Wednesday, October 16. The bulletin: 489 deaths. The peak, though the Bureau of Health would not know it for several days. The figure appeared in the Ledger with a qualification: the numbers included only deaths reported to the Bureau of Health. The qualification was necessary. The city’s death registration system, which depended on physicians filing certificates with the local registrar, was now days behind. Bodies were being buried without certificates, or with certificates filed retrospectively, or not filed at all. The true death toll for October 16 would later be estimated at 700 or more. The 489 recorded deaths were a measure of administrative capacity, not mortality.

The sensory world of the city had transformed. The ambulance bells, which had been occasional in September, now rang continuously. The Philadelphia Hospital, the city’s public institution for the indigent, reported that its morgue, designed for forty bodies, held 140. The hospital’s superintendent, Dr. Joseph Leidy, telephoned the Bureau of Health to request assistance with burial. The Bureau had no assistance to send. The city’s regular force of undertakers, approximately 200, was overwhelmed. Coffins, which had been in short supply since October 5, were now unobtainable through ordinary commerce. The C.J. McElwee Company, the largest manufacturer in the region, was producing 300 caskets daily and selling them directly to the city at cost. Still insufficient.

The sound of the drays began. Horse-drawn wagons, borrowed from transfer companies and furniture movers, carried loads of plain wooden boxes to the cemeteries. The Bulletin reported on October 16 that the city’s cemeteries had received 1, 200 burials in the preceding seventy-two hours. The report appeared on page six. The same edition carried, on page one, a photograph of General Pershing decorating soldiers in France.

The confidence that had sustained the city through September was now broken. That confidence had operated on a simple mechanism: official optimism suppressed alarming reports, which made optimism seem justified, which encouraged further suppression. The mechanism required that the alarming reports remain deniable. On October 16, they became undeniable. The 489 deaths could not be minimized into a seasonal catarrh. The sight of drays on city streets could not be described as normal. The silence of the newspapers was no longer a strategy of reassurance but a failure of comprehension. They had no form for this story.

Thursday, October 17. The bulletin: 447 deaths. The decline from the peak was slight, possibly within the margin of reporting error, but it would be seized upon in subsequent accounts as evidence of improvement. The Inquirer headlined that the epidemic showed decided decline, a judgment based on a single day’s figures against a three-week rise. The paper’s optimism was not shared by the institutions actually managing the crisis. The Board of Health, meeting in emergency session, authorized the requisition of additional cemetery land and the impressment of labor for grave-digging. The seminarians of St. Charles Borromeo, who had been digging trenches in Holy Cross Cemetery since October 10, were now joined by crews from the city prison. The prisoners worked in chains, under guard, digging trenches six feet wide and forty feet long, into which coffins were placed without individual graves.

The visual evidence of this labor was visible from the street. Holy Cross Cemetery, in the city’s southwest, occupied high ground visible to passengers on the Baltimore and Ohio Railroad. Travelers on the 4:15 train to Washington reported seeing long trenches and continuous funerals through the cemetery’s perimeter fence. These reports appeared in out-of-town papers before they appeared in Philadelphia. The Washington Post ran a dispatch on October 18 describing mass burials in Philadelphia; the Inquirer did not confirm this description until October 21, and then only to deny that the burials were indiscriminate.

The denial was technically accurate. The trenches were organized by date of death, and records were kept of each placement. But the visual impression, of coffins stacked in rows without the spacing and markers of individual graves, was not mistaken. The city had moved from individual burial to collective disposal, and this transformation was visible to anyone who passed the cemetery gates.

Friday, October 18. The bulletin: 389 deaths. The Ledger placed this figure on page four, with a subheading that the worst was believed to be over. The belief was premature. The mortality curve would remain elevated for another week, and the secondary deaths—from pneumonia, from heart failure, from the exhaustion of convalescence—would continue into November. But the psychological turning point had passed. The city had survived the week of four thousand dead.

The count, reconstructed from the daily bulletins, was as follows: October 12, 289; October 13, 289; October 14, 376; October 15, 428; October 16, 489; October 17, 447; October 18, 389. The total for the seven days was 2, 707 reported deaths, with the actual toll certainly higher. The week of four thousand dead was a figure arrived at by including the surrounding days—October 10 through October 20—in which the daily toll exceeded 200. This broader period recorded approximately 4, 500 deaths, a rate that translated to an annual mortality of 1, 200 per 100, 000 population, roughly six times the city’s normal rate.

The transformation of the city’s social fabric was now complete. The factories that had closed for lack of workers would reopen slowly, with new crews and reduced output. The theaters that had darkened would remain dark, by official order, until October 26. The schools, closed since October 3, would not reopen until November. The churches, which had held services on October 13 against the Board of Health’s recommendation, would cancel public worship by October 20. The archdiocese of Philadelphia, which had resisted closure as an infringement on religious liberty, announced on October 19 that all Sunday masses were suspended until further notice.

The further notice would come in three weeks. The epidemic would decline through November, not to zero but to a manageable baseline. The city would record its final influenza death of 1918 on November 11, the day of the armistice. The symmetry was noted by no one at the time. The armistice celebrations, which brought crowds back to Broad Street on November 11, were covered by the newspapers with the same triumphal vocabulary as the Liberty Loan parade. The epidemic was not mentioned in the armistice editions.

—-

The Bureau of Health’s bulletins for the week of October 12–18 survive in the municipal archives, bound in a leather volume with the mortality returns for 1918. The typed sheets are fragile, the carbon copies faded. The handwriting of the clerks who compiled the figures is visible in the margins, checking totals, carrying numbers forward. The volume contains no commentary. The clerks who made these entries were not asked to interpret their meaning, only to record it.

The newspapers of the same week survive in the library of the Historical Society of Pennsylvania, on microfilm and in bound volumes. The contrast between the bulletins and the papers is instructive. Where the bulletins show a continuous numerical ascent, the papers show a discontinuous attention, a capacity to see and not see the same facts. The 428 deaths of October 15 appear in the Ledger as a figure without context. The 489 deaths of October 16 appear as a decided decline in the Inquirer of October 17, based on a misreading of the previous day’s partial report. The correction, if it was made, appeared in smaller type.

Not conspiracy. The failure of a system designed for other purposes. The city’s newspapers were organized to report war news, commerce, society, and sport. They had no department for epidemic, no form for mass death. The wire services, on which they depended for national and international coverage, were similarly silent. The Associated Press carried no dispatches on the Philadelphia epidemic during the week of October 12–18. The national papers—the New York Times, the Washington Post—carried brief notices of severe conditions in eastern cities, without specificity. The epidemic was happening everywhere and therefore nowhere, a universal condition that exceeded the particularizing logic of news.

The sensory record, what the city actually experienced, survives in fragments. The silence of the factories is documented in the absenteeism reports of the Ordnance Department. The empty streetcars are recorded in the Philadelphia Rapid Transit Company’s monthly statement. The ambulance bells are mentioned in a memoir by Dr. George A. Soper, the public health engineer who visited Philadelphia in late October to assess the situation. Soper described a city of frightened people, a phrase that appeared in his official report but not in any Philadelphia paper.

The drays and the trench graves are visible in the photographs of the City Archives, images made by an unknown photographer who walked through Holy Cross Cemetery on October 17 or 18. The photographs show wooden boxes in rows, earth piled beside open trenches, men in civilian clothes and clerical collars working with shovels. The photographer’s name is not recorded. The photographs were not published in 1918.

The volunteer network, which had sustained the city through the first phase of the crisis, was now itself depleted. The Visiting Nurse Society reported that 34 of its 125 nurses had contracted influenza, and that 4 had died. The Emergency Aid, which had answered 800 calls on October 14, could answer 200 by October 18. The volunteers were not immune. The seminarians digging at Holy Cross worked in rotation, with prayers between shifts. Two of them died in the third week of October, their names added to the cemetery’s records as workers rather than as interments.

The city’s institutions, which had failed to prevent the epidemic, now failed to manage its documentation. The death certificates that were filed in late October were often incomplete, with influenza or pneumonia given as cause without specification of type. The Bureau of Health’s statisticians, attempting to compile accurate totals for the year, found discrepancies of 15 percent between their own bulletins and the registrar’s final count. The discrepancy was never fully resolved. The official death toll for Philadelphia in 1918 from influenza and pneumonia was eventually fixed at 12, 191, but this figure included deaths through December and was known to undercount the October peak.

The week of four thousand dead was thus a statistical construction as well as a physical experience. Later historians, working from fragmentary records, gave it shape; the people who lived through it experienced not a week but a continuous present of uncertainty, in which each day’s bulletin might be the worst or might be superseded. The peak was visible only in retrospect. On October 16, 1918, no one knew that the 489 deaths were the maximum. On October 17, the slight decline might have been a fluctuation. On October 18, the Ledger’s confidence in decided decline was as unjustified as its previous silence had been complicit.

The verdict that normalcy was irrevocably broken arrived not with the peak but with the recognition that there would be no immediate return. The theaters did not reopen on October 19. The schools did not resume. The churches, having resisted closure, now accepted it. The volunteer network, having stretched to its limit, now required reinforcement from conscripted labor.

The city had passed from a condition of emergency managed by civic spirit to a condition of emergency managed by institutional improvisation. The improvisation would continue for another month, through the decline of the epidemic and the return of something resembling ordinary life.

But the ordinary life that returned was not the same as the life that had preceded the parade. The confidence that the city’s systems could manage any challenge, that the protocol of normalcy was itself a form of protection, had been buried in the trenches of Holy Cross Cemetery, in coffins that arrived by dray and were placed in rows without ceremony.

The Bureau of Health’s clerk who typed the bulletin for October 19, 1918, recorded 363 deaths. He added a note in the margin: Continuing improvement expected. The expectation was not fulfilled in the manner anticipated. The improvement was real but slow, and it carried with it the knowledge of what had been lost. The city would not forget, but it would also not remember in any organized way. No monument was erected to the dead of October 1918. The reorganization of the Department of Public Health, which followed in 1919, was described in administrative language that made no reference to the specific failures of the preceding autumn. The Liberty Loan parade, which had seeded the disaster, was remembered as a patriotic success. The epidemic was remembered, when it was remembered at all, as a natural catastrophe, an act of God or of virus, for which no human agency could be blamed.

The final silence, the silence that followed the silence of the newspapers. The city had learned to live with death rates that would have been inconceivable in September, and it had learned to look away from the learning. The week of four thousand dead established that the pathogen could triumph over any social fabric, that the rituals of normalcy could accelerate rather than prevent disaster, and that the institutions charged with protection could fail without acknowledging their failure. These lessons were not recorded in the official minutes. They were written in the trenches of the cemeteries, in the empty seats of the streetcars, in the pages of newspapers that reported everything except what was happening.