The Shadow in the Ranks

The posters on the walls of Philadelphia showed a doughboy fixing bayonets, his face set toward an unseen enemy. The parade was six days away. At the Navy Yard, where the Delaware River widened into its tidal basin, a medical corpsman made an entry in a leather-bound sickbay log. The date was September 7, 1918. The entry gave a name, a rating, a temperature of 103 degrees, and a diagnosis: “epidemic catarrh.” The hand that wrote it was steady, practiced, the hand of a man who expected to make many such entries before his watch ended.

The term placed the illness in a familiar category. Catarrh—respiratory inflammation, common in crowded barracks, usually self-limiting, rarely fatal in young men. The modifier “epidemic” acknowledged that multiple cases were appearing, but the protocol did not require alarm. The Navy’s medical service had managed influenza outbreaks before. The war had already produced several waves of illness in training camps and transport ships, episodes that disrupted schedules but yielded to standard measures: isolation of the sick, disinfection of quarters, temporary quarantine of exposed units. The logbook entry was routine.

By September 10, the logbook required additional pages. The same diagnosis appeared with variations: “Spanish influenza,” “influenza,” “grippe.” The symptoms followed a pattern: sudden onset, high fever, severe headache, pain in the back and limbs, cough, and in some cases, a dusky discoloration of the face that suggested oxygen starvation. The physicians noted these signs without comment. They had no laboratory confirmation of the pathogen. The influenza virus would not be isolated until 1933. In 1918, diagnosis remained a clinical art, a matter of recognizing patterns in symptoms and comparing them to accumulated experience.

The sickbay occupied a long, low building of brick and timber near the center of the Yard. Its wards were arranged in the pavilion style then favored for contagious disease: high ceilings, cross-ventilation, beds spaced to prevent droplet transmission. These designs had been developed for tuberculosis and diphtheria, diseases whose mechanisms of spread were understood. The influenza virus, transmitted through fine aerosols and capable of surviving on surfaces, exploited the gaps in this knowledge. Men recovered in wards where new patients arrived hourly. The ventilation that was meant to disperse contagion may have helped distribute it. The spacing of beds meant little when nurses and orderlies moved between patients, when the same thermometer was used for multiple readings, when the air itself carried infectious particles.

The first deaths occurred on September 11 and 12. They were recorded in a separate ledger, the mortality register, with the same clinical detachment: name, rank, age, date of death, cause. “Bronchopneumonia following influenza.” “Lobar pneumonia.” “Acute pulmonary edema.” The causes were accurate as far as they went. The influenza virus stripped the respiratory epithelium, clearing the way for bacterial invasion of the lung. The resulting pneumonia, often hemorrhagic and frequently bilateral, killed with a speed that shocked experienced clinicians. A man could be ambulatory at morning sick call, febrile by noon, cyanotic by evening, dead by morning. The physicians had seen pneumonia before, but not in such concentration, not in such young and previously healthy bodies, not with such resistance to treatment. The standard therapies—digitalis for cardiac support, oxygen inhalation, stimulants—seemed to have little effect on the downward course.

The Philadelphia Navy Yard sat at the southern edge of the city, a 900-acre complex of shipways, foundries, barracks, and hospitals that had swallowed the earlier League Island and transformed it into the engine of American naval expansion. In September 1918, some 45, 000 sailors and civilian workers passed through its gates daily, a population larger than most American towns, compressed onto a finger of land that pointed toward the sea lanes of the war. The Yard’s medical officers faced a dilemma that would recur throughout the epidemic. Their duty was to maintain the fighting strength of the fleet. Every sailor hospitalized was a sailor unavailable for duty. Every ship delayed in port was a ship not carrying troops to France, not escorting convoys against U-boats, not projecting American power into the European war.

The war had created a hierarchy of urgency in which medical considerations competed with operational requirements for attention and resources. The Yard’s command structure operated under direct orders from the Bureau of Navigation in Washington, which operated under the direction of the Secretary of the Navy, which operated within the coordinated framework of the Council of National Defense. The first telegrams to the city’s Bureau of Health were sent on September 14. They were routine notifications, required by the city’s sanitary code for certain communicable diseases. The telegrams identified the Navy Yard as the source and requested that the city health authorities “take cognizance” of the outbreak.

The phrasing was significant. The Navy maintained its own medical service, its own hospitals, its own quarantine facilities. The request for city cognizance was a formality, a courtesy between jurisdictions. It did not constitute an alarm or a request for assistance. The city Bureau of Health, located in the Municipal Building at Fifteenth and Chestnut Streets, received the telegrams and filed them. The Bureau’s director, Dr. Wilmer Krusen, was a physician of established reputation, trained in Philadelphia’s medical schools, experienced in public health administration, and committed to the progressive vision of sanitary science that had transformed American cities in the preceding decades. Krusen had taken office in 1917, inheriting a department that had managed typhoid outbreaks, supervised milk inspection, and coordinated the city’s response to the national influenza epidemic of 1889-1890. He understood the language of official communication. The Navy’s telegrams did not require immediate action. They requested cognizance, not intervention.

Dr. Paul Lewis received his first samples from the Navy Yard on September 15. Lewis occupied a position at the intersection of Philadelphia’s scientific and medical establishments: professor of pathology at the University of Pennsylvania, director of the Henry Phipps Institute for the Study, Treatment and Prevention of Tuberculosis, consultant to the city’s hospitals and health department. The Phipps Institute, located at Ninth and Lombard Streets, was a model of the new scientific medicine, equipped with laboratories for bacteriological research, X-ray apparatus, and facilities for animal experimentation. Lewis had trained with Simon Flexner at the Rockefeller Institute and had already made contributions to the understanding of poliovirus and tumor pathology. He was, by 1918, one of the most respected bacteriologists in the United States, a man whose judgment would carry weight in any deliberation about the nature of an infectious outbreak.

The samples arrived in glass tubes, packed in ice, accompanied by clinical notes describing the course of illness in the donor patients. Lewis and his assistants set to work with the methods then available: culture of bacteria from blood and sputum, microscopic examination of stained tissues, inoculation of experimental animals to test pathogenicity. The bacteriological culture yielded mixed results. The lungs of fatal cases showed a variety of organisms—pneumococci, streptococci, staphylococci, the rod-shaped bacteria of influenza that we now recognize as Haemophilus influenzae. This polymicrobial picture was consistent with secondary bacterial invasion, but it did not identify a primary pathogen. The clinical pattern, however, was distinctive. The rapid progression, the hemorrhagic pneumonia, the unusual mortality in young adults—these features suggested something beyond the ordinary seasonal influenza that Lewis had encountered in his practice.

Lewis kept a laboratory notebook in which he recorded his observations, his hypotheses, and his experimental protocols. The entries for September 15-20 show a mind working through possibilities, testing each against the evidence.

He noted severe toxemia, pulmonary lesions characteristic of fulminating infection, no single organism consistently isolated, and the possibility that a filterable virus must be considered. The concept of a filterable virus—invisible to microscopy, passing through porcelain filters that retained bacteria—was still controversial in 1918. The existence of such agents had been demonstrated for diseases like yellow fever and rabies, but their nature remained mysterious. Lewis’s willingness to consider a viral etiology placed him at the forefront of contemporary research.

It also placed him in a position of uncomfortable knowledge. If the outbreak was caused by a novel virus, there was no specific treatment. The antiserum therapies developed for bacterial infections would be useless. Prevention would depend on isolation and quarantine measures of a scope that would disrupt the war effort.

The virus was already moving beyond the Navy Yard’s boundaries. The southern wards of Philadelphia, the neighborhoods of Southwark and Moyamensing that clustered around the Yard’s gates, began to report unusual illness in the second week of September. The tenement houses of these districts housed the civilian workers who walked to the shipways each morning, the laundresses who washed naval uniforms, the shopkeepers who served the Yard’s population. The housing stock was old, much of it dating to the early nineteenth century, constructed without regard to the sanitary reforms that had transformed other parts of the city. Water closets were often shared between multiple families. Ventilation was poor. The density of population—fifty, sixty, eighty persons per acre in some blocks—created conditions ideal for respiratory transmission.

The first civilian cases appeared in households with direct connection to the Yard: a riveter’s wife, a machinist’s children, a café owner who served the noon trade of naval officers.

The neighborhood physicians, general practitioners who managed the full range of urban illness, recognized the severity of what they were seeing. Dr. Howard Anders, a physician with practice in the Twelfth Ward, later recalled a sudden influx of cases with high fever and prostration, unlike anything in his previous experience. The recall was made in retrospect, shaped by the knowledge of what followed. In the moment, individual physicians faced individual patients, each case demanding its own management. There was no mechanism for aggregating these observations into a picture of epidemic spread. The city’s vital statistics system recorded deaths, not cases. The reporting of communicable disease depended on physicians’ compliance with notification requirements, compliance that was incomplete even in the best-managed jurisdictions. The Bureau of Health received scattered reports, filed them, and waited for a pattern to emerge.

Krusen maintained his public position of confidence. On September 18, he issued a bulletin to the city’s newspapers stating that no contagious disease of any importance existed in Philadelphia at that time.

The statement was technically defensible. The Navy Yard outbreak was a federal jurisdiction. The civilian cases were not yet numerous enough to meet the statistical threshold for epidemic declaration. The bulletin reflected the standard protocol of public health administration: avoid alarm, maintain confidence, intervene with precision rather than precipitation.

Krusen was not a negligent official. He was a man trained in a system that valued measured response, that had learned from the excesses of nineteenth-century quarantine, that prioritized the maintenance of social and economic function. The war had intensified these priorities. Civilian morale was a strategic resource. The Liberty Loan drive, with its $259 million quota for Philadelphia, required public gatherings, patriotic display, and the demonstration of collective commitment. Any suggestion of epidemic danger would threaten these operations.

The Navy Yard’s command faced its own pressures. The command structure received daily reports from medical officers, reports that showed continued increase in cases and deaths. It also received operational requirements from Washington: the transport Leviathan was due to sail with troops on September 26; the destroyer squadron required immediate overhaul; the new battleship Pennsylvania needed crew complement. The medical reports competed with these requirements for attention. The standard response to infectious outbreak in a naval unit was isolation and quarantine, but quarantine of the entire Yard would paralyze the Atlantic fleet’s logistical system. The compromise solution, adopted around September 16, was selective quarantine: isolation of affected barracks, restriction of liberty for exposed units, continuation of essential operations. This was the protocol of normalcy applied to military medicine: maintain the appearance of function while managing the reality of disease.

The telegrams between the Yard and the Bureau of Health continued through the third week of September. They grew more detailed in their descriptions of cases, more urgent in their requests for city cooperation. On September 20, the Navy medical officer reported approximately 500 cases under treatment, with 20 deaths to date. The numbers were approximate; precise accounting was difficult in the chaos of expanding wards. The city response remained procedural. Krusen assigned an inspector to coordinate with Navy medical authorities. He requested daily reports. He prepared the forms that would be needed if the outbreak spread to civilian population. He did not, in the surviving records, propose any measure that would interfere with the Yard’s operations or the city’s scheduled events.

Dr. Lewis’s laboratory work intensified. By September 22, he had completed initial passages in experimental animals, attempting to transmit the disease to monkeys and ferrets. The results were inconclusive. Some animals showed signs of illness; others remained unaffected. The variability suggested either a filterable virus of low pathogenicity in non-human species, or a complex etiology involving multiple factors. Lewis prepared a preliminary report for the Philadelphia County Medical Society, scheduled to meet on September 24. The report was circumspect in its conclusions, emphasizing the severity of the clinical picture while acknowledging uncertainty about the causative agent. The present outbreak, he wrote, presented features that distinguished it from ordinary epidemic influenza, though its ultimate nature remained to be determined. The language was that of scientific caution, appropriate to incomplete data. It was also language that could be read as reassurance: the distinguished Professor Lewis had examined the cases and found nothing definitively alarming.

The Medical Society meeting was attended by representatives of the city’s hospital staffs, medical school faculties, and public health administration. Lewis’s presentation was followed by discussion, recorded in the Society’s minutes. Several physicians reported similar cases from their own practices. The consensus that emerged was that Philadelphia faced an unusual but manageable outbreak of respiratory disease, that standard preventive measures should be applied, and that there was no cause for public alarm. The minutes do not record any dissent from this consensus. The physicians in attendance were not stupid or careless men. They were practitioners of a medicine that had achieved remarkable successes against infectious disease through sanitation, immunization, and clinical care. They had no framework for recognizing a novel pandemic virus, no experience with an agent that killed through cytokine storm and secondary bacterial invasion, no precedent for the speed of transmission that would soon overwhelm their institutions.

The Liberty Loan parade remained on the calendar for September 28. The posters on the city’s buildings showed the familiar imagery: the doughboy advancing, the flag flying, the citizen’s duty to lend. The organizational machinery of the loan drive, directed from the Treasury Department through regional committees, had planned the parade as the climax of Philadelphia’s subscription campaign. The route was set: from the Navy Yard, up Broad Street, past City Hall, to the Bellevue-Stratford Hotel where officials would review the marchers. The participation of the Yard’s personnel was considered essential: sailors in formation, bands playing, the demonstration of naval strength that would inspire civilian commitment. The medical officers who knew the condition of their commands did not, in the surviving record, protest this plan. Their duty was to restore men to duty, not to dictate operational decisions. The protocol of normalcy required that medical concerns be subordinated to strategic requirements.

By September 25, the Yard’s medical facilities were approaching capacity. The sickbay wards had been expanded into temporary structures, tents and converted warehouses, but the flow of new cases exceeded the available beds. The mortality rate, calculated crudely from admissions and deaths, suggested something between 2 and 4 percent of recognized cases—lower than the catastrophic figures that would emerge in the general population, but high enough to alarm any physician. The Navy Surgeon General’s office in Washington had been notified and had dispatched additional medical personnel, including specialists in infectious disease and additional nursing staff. These reinforcements began arriving on September 26, too late to change the trajectory of the Yard outbreak, but in time to witness its peak.

Lewis received additional samples on September 26, these from civilian cases in the Southwark district. The bacteriological findings were consistent with those from the Navy: mixed bacterial flora, no single pathogen dominant, clinical pattern suggesting primary viral infection with secondary bacterial complications. He prepared additional slides, made additional cultures, continued the animal passages that might eventually yield a reproducible model of the disease. The work was meticulous and slow. Science, in September 1918, had no faster method for characterizing an unknown pathogen. The very precision of Lewis’s methods—his insistence on controlled conditions, reproducible results, conservative interpretation—meant that his findings could not keep pace with the epidemic’s spread.

The city Bureau of Health received a final pre-parade report from the Navy Yard on September 27. The report stated that the outbreak was now believed to be under control, with declining numbers of new admissions. This was accurate as a description of the Yard’s immediate situation: the peak of the naval outbreak had passed, not because of any intervention, but because the virus had exhausted the susceptible population in the compressed environment of the barracks. The statement was misleading as a forecast of the city’s situation. The virus had already seeded the civilian population. The incubation period—typically two to three days—meant that hundreds of infectious individuals would be present at the parade, shedding virus in the dense crowds that gathered along Broad Street. The Bureau of Health filed the report. Krusen issued no warning. The posters remained on the walls.

Dr. Paul Lewis’s private laboratory confirmation of a virulent strain came on September 28, the morning of the parade. His notebook records the completion of a series of passages in which pooled material from fatal cases had produced consistent pathological changes in experimental animals. The filterable nature of the causative agent, he wrote, was now established beyond reasonable doubt. The virulence of the strain, as measured by mortality in inoculated ferrets, exceeded that of any influenza virus previously studied. The finding was definitive in scientific terms. It was also, on that morning, irrelevant to public policy. The parade was assembling at the Navy Yard. The crowds were gathering. The city’s public bulletins for that week described the situation as under control.