In the winter of 1906, a rented summer house on Long Island sat empty: nobody could explain why six of its eleven occupants had caught typhoid the previous August. The first investigators had analyzed the water and found nothing wrong, and traced dye from the cesspools to confirm none of it reached the well. The owner, afraid she would never rent the place again, hired George Soper, a sanitary engineer who had spent years chasing typhoid outbreaks for New York City and State, to look again.
Soper agreed the water was clean, and dropped a theory of his own just as fast: a neighbor's shellfish, ruled out once he found nobody in the house had eaten any for six weeks. Only then did he ask for a calendar. The cook had started August 4, three weeks before the first of six patients fell ill, the rest within seven days. That alone proved nothing, so Soper traced her backward through ten years of employers. In 1902, at a rented house in Maine, seven of nine people fell ill within weeks of her arrival. In 1904, on a Long Island estate, four people in the servants' quarters fell ill three weeks after the family moved there, though she'd worked for them nine months already. Eight families in ten years; typhoid struck seven, not always quickly — at one house the gap ran three years. Nothing else connected the outbreaks. The cook did.
When Soper found her, cooking in a Park Avenue house where a laundress had just been hospitalized and the family's daughter lay dying, he asked for her urine, feces, and blood. She picked up a carving fork and came after him; he left fast. It took the health department, Dr. S. Josephine Baker, three policemen, and an ambulance to take her, after officers found her behind ash cans in a neighbor's yard. Soper said he hadn't needed a laboratory at all: the dates and addresses had already made his case. The culture, once it grew, only agreed — testing her stool three times a week for eight months, the lab found Bacillus typhosus in it almost every time.
Soper did not invent the idea of a healthy carrier. German researchers had already described people who carried typhoid bacilli without symptoms, in papers Soper said he mostly missed until a colleague pointed him to them after the case was closed. What he produced was the first such case documented in America, built before any laboratory was involved: a stack of records, dates of arrival set against dates of illness, matched across seven households, naming a cause before a culture existed. A record like that can still name a cause months before a microscope does.
New York held her nearly three years, then released her on her word she would never cook again. She broke it under a new name; in 1915 a Manhattan hospital's kitchen produced an outbreak of more than twenty cases traced to her too. She spent the next twenty-three years confined on North Brother Island and died there in 1938. Soper's own tally, fifty-three infected and three dead, counts only what investigators could trace; her five free years were never fully retraced, and she had probably infected others nobody connected to her. The city caught the dates, eventually. It never caught up with everywhere she had cooked.