The Navy sprayed San Francisco to see what would happen.

September 1950.

The war had been over for five years, and the Army had not stopped thinking about the next one. At Camp Detrick in Maryland, the Chemical Corps ran a biological warfare program under heavy secrecy, and it had a question a laboratory could not answer. An enemy would not need a bomber to reach an American city. A ship could sit offshore, upwind, and let a cloud go, and the wind would carry it in over everyone. What the program wanted to know was how far such a cloud would drift, how long it would stay in the air, and how many people would breathe it before anyone noticed anything at all.

That was not something a wind tunnel could tell them. The only way to get a real answer was to try it over a real city, with real weather and real people going about their day, and then measure where it landed.

San Francisco suited the purpose. It sat at the end of a peninsula with the ocean on one side and the bay on the other, and in the afternoons the wind came in off the Pacific and pushed across the city through the gap at the Golden Gate. The fog came with it, on a rhythm the whole region knew. A cloud released into that air would have helped move inland.

The simulant the program used had to do two things. It had to behave in the air roughly the way a real biological agent would, drifting and settling the same way, and it had to be something the testers could find again afterward so they could measure where it had gone. For that second job they used a bacterium called Serratia marcescens. It had a convenient property. When it grows on a culture plate, it forms colonies that are red. Collect air samples across the city, grow what you catch, and the red dots tell you the cloud reached that corner. It was treated at the time as a harmless organism, the kind that turned up on old bread and was used in classrooms precisely because it was thought to be safe.

Over several days in late September 1950, a Navy vessel offshore released the bacteria into the air, along with another test organism, and let the wind do the rest. The spraying ran for the better part of a week. Monitoring stations spread around the Bay Area then counted what came down. The results told the program what it wanted to know: the cloud had spread across a wide area of the city, and by the testers’ own reckoning, nearly everyone in San Francisco would have breathed in some of it. Hundreds of thousands of people. None of them knew a test was underway, because the point of the test was partly to see what happened in a population that did not know.

A short distance from the waterfront stood Stanford University Hospital. In the weeks after the spraying, the doctors there began seeing something they had not seen before. Patients were turning up with infections caused by Serratia marcescens, the same red bacterium, mostly in the urinary tract, several of them in people who had recently had procedures done. This was strange on its face, because Serratia was supposed to be the organism that did not make people sick. The hospital had no record of it causing trouble like this.

One of the patients was an older man named Edward Nevin. He had been admitted after surgery and was expected to recover. Instead, he developed a Serratia infection that settled on a heart valve, the kind of infection that, in 1950, there was very little medicine could do about. He got worse over a period of weeks, and he died.

The doctors did what careful doctors do with something they cannot explain. They treated the patients, they tracked the cluster, and they wrote it up and published it in the medical literature, because a sudden run of infections from an organism nobody expected to cause them was worth putting on the record for other physicians to see. What they could not put in the report was the cause, because they did not know it, and there was no way for them to know it. The thing that would have explained the whole cluster was classified, held inside a program they had never heard of, in a building on the other side of the country.

The answer stayed inside that program for a quarter of a century. Through the 1950s and into the 1960s, the open-air testing continued in other places, over other cities and transit systems and stretches of the country, part of a large effort to understand how an attack would spread and how exposed the United States was to one. The people underneath the clouds were not told then either.

What finally brought it out was Washington. In the middle of the 1970s, the government’s intelligence and defense programs came under a wave of congressional investigation, and the biological testing surfaced along with everything else. In 1977, a Senate subcommittee held hearings on the Army’s open-air tests, and Army witnesses laid out the scope of the program for the record. Among the operations described was the one conducted over San Francisco in 1950.

The Nevin family was reading the newspapers. Edward Nevin had a grandson, also named Edward Nevin, who was a lawyer, and when the hearings connected the 1950 spraying to the same bacterium and the same window of time as his grandfather’s death, he brought a case against the United States.

The lawsuit went to federal court in San Francisco. The family argued that the government had released bacteria over a city full of people, that one of those people was Edward Nevin, and that he had died of an infection caused by the very organism the government had sprayed. The government argued, among other things, that the decision to run the test was the kind of high-level policy choice that the law shields from lawsuits, under the discretionary function exception to the Federal Tort Claims Act. The court agreed with the government. The claim was dismissed, the dismissal was upheld on appeal, and the Supreme Court declined to take the case. The question of whether the government had been careless was never reached, because the court held it was not a question the lawsuit was allowed to ask.

No one was charged with anything. The testing program had already been wound down by the time the hearings exposed it. The scientific debate over whether Serratia marcescens is truly harmless shifted after San Francisco, and the organism is now recognized as one that can and does cause serious infection, particularly in hospitals and in people who are already sick. The classroom demonstrations with the pretty red bacterium largely stopped.

The published account of the Stanford outbreak is still in the medical literature where the doctors left it: a careful description of eleven patients and the organism that infected them, written by physicians who did their jobs and never learned why the patients came.