September 1950: a week of fog in a city that didn't know about the test.

September 20, 1950.

September is San Francisco's warm month. The summer fog that pours in through the Golden Gate in June and July thins out, the wind drops, and the city gets its best weather while the rest of the country heads into fall. In September 1950, the Korean War was less than three months old. The shipyards and Navy installations around the bay were busy again. Hundreds of thousands of people lived on the hills of the peninsula, packed close together, with the Pacific on one side and the bay on the other.

The bacterium at the center of this story lived almost everywhere: soil, water, damp corners of kitchens. It was called Serratia marcescens. Its useful trait was color. Grown on a laboratory plate, it formed colonies of bright red, easy to pick out among other microbes. That made it valuable to anyone who wanted to follow a germ from one place to another and find out where it ended up. In 1950, the researchers who used it widely considered it harmless.

After the Second World War, the United States built a biological weapons program. Its center was Camp Detrick in Maryland, run by the Army Chemical Corps. The planners had a question they could not answer in a laboratory. If an enemy ship sailed up to an American coastal city and released a cloud of germs into the wind, how far would the cloud travel? How much of it would people breathe? Would it work at all?

To find out, they needed a real city, real weather, and real lungs. The Navy and the Army planned a test. It was called Operation Sea-Spray.

On September 20, 1950, a Navy vessel moved into position off the coast near the Golden Gate. Equipment on board sprayed a fine mist into the onshore wind. The mist carried two kinds of bacteria. One was Serratia marcescens. The other was Bacillus globigii, another organism the scientists treated as harmless and used because it was easy to identify. The wind carried both toward the city.

Over the next week, the ship made a series of these runs, each one a simulated attack. On shore, sampling stations had been set up across San Francisco and around the bay. They pulled in air and collected whatever it held. Back in the laboratory, the samples were grown out, and the red colonies were counted.

The counts told the planners what they wanted to know. The clouds had drifted across the city. They had reached neighborhoods miles from the water. By the testers' own estimates, almost everyone in San Francisco had breathed in some of the bacteria. A single ship, sitting offshore and doing nothing more complicated than spraying, could expose an entire American city. In the program's terms, the test was a success.

No public announcement was made. The people of San Francisco went to work, rode the cable cars, hung laundry, and opened their windows on warm afternoons. The results went into a classified report.

At that time, Stanford University's hospital was not on the Palo Alto campus. It was in San Francisco. Near the end of September, doctors there found Serratia marcescens in a patient's urine. That was strange. The organism was almost never seen as a cause of human infection. Then they found it in another patient, and another.

Over the following months, the hospital recorded about a dozen of these infections. Most were in the urinary tract. Many of the patients had recently gone through procedures that involved the bladder or the urinary tract. The doctors had not seen anything like it, and they could not explain where the organism had come from.

One of the patients was Edward Nevin, a retired pipefitter in his seventies. He had come into the hospital for prostate surgery. It was supposed to be the kind of operation a man recovers from and goes home. After the procedure, he developed a Serratia infection. It spread through his bloodstream to his heart and settled on a heart valve. Weeks after he was admitted, Edward Nevin died.

His family buried him. They had been told he died of an infection after surgery, which was true as far as anyone at the hospital knew.

Doctors at the hospital wrote up the cluster of cases and published it in a medical journal in 1951. They described an unusual outbreak of an organism that rarely made people sick. They had no way to link it to a Navy ship that had been offshore a few days before the first case appeared. The test was classified.

The testing program went on. Through the 1950s and 1960s, the Army ran open-air trials in and around American cities and transportation systems, studying how particles moved. In 1966, Army teams went into the New York City subway and dropped light bulbs filled with bacteria onto the tracks, then measured how far the trains carried them through the tunnels. The bacteria in those tests were again ones the Army considered harmless.

In November 1969 President Richard Nixon announced that the United States was giving up offensive biological weapons. The program was shut down. Its records stayed classified.

The story came out in pieces. In December 1976, the Long Island newspaper Newsday reported on the Army's past open-air tests, including the spraying of San Francisco. Other papers followed. In San Francisco, the Nevin family read the coverage. Edward Nevin's grandson was a lawyer. He started looking into his grandfather's death and the dates.

In 1977, a Senate subcommittee chaired by Edward Kennedy held hearings on the Army's biological testing. The Army disclosed that between 1949 and 1969 it had conducted open-air tests in hundreds of locations across the country, many of them populated. Sea-Spray was one of them. Army witnesses maintained that the bacteria had been believed harmless at the time and that the tests were a reasonable way to measure how vulnerable the country was.

Across the medical world, the view of Serratia marcescens had changed. In the decades since 1950, it had turned up again and again in hospitals, infecting the urinary tracts, lungs, blood, and heart valves of patients who were old, sick, recovering from surgery, or carrying catheters. The organism the testers had picked for being safe had become a known hospital pathogen.

The Nevin family sued the United States government. The case went to trial in federal court in San Francisco in 1981. The family argued that the bacteria sprayed from the Navy ship had reached the hospital and killed Edward Nevin. The government argued that the bacteria were believed harmless, that the Army's decision to run the test was protected from lawsuits, and that nobody could prove the strain in Nevin's blood was the strain from the ship. Samples from the 1950 patients had not been kept in a way that allowed that comparison.

The judge ruled for the government.

The family appealed and lost. They asked the Supreme Court to take the case, and it declined.

No criminal charges were brought against anyone involved in Operation Sea-Spray.

The Nevin family received no damages. The government never accepted that the test had caused Edward Nevin's death, and the court never found that it had.

The 1977 Senate hearings put the existence of the open-air testing program on the public record. The Army's own accounting showed that San Francisco was one of hundreds of places it had used for tests over twenty years.

The United States signed the Biological Weapons Convention, which bans developing and stockpiling biological weapons. Camp Detrick, later Fort Detrick, became a center for defensive biological research.

Stanford's hospital moved to the Palo Alto campus at the end of the 1950s. The old building in San Francisco was taken over by another medical center.

The Golden Gate still fills with fog every summer, and the wind still blows in off the Pacific across the city, the way the testers counted on.

Serratia marcescens is now on the list of organisms hospitals watch for in their patients, especially the elderly and those recovering from surgery.