Disturbing History - Operation Sea Spray

Episode Date: August 19, 2026

In September 1950, a United States Navy vessel spent eight days off the San Francisco coastline spraying clouds of live bacteria into the wind, and nobody on shore was told.The operation was called Sea-Spray, and it was designed to answer a question that terrified early Cold War planners: could an enemy attack an American city with a biological weapon delivered from a small boat, invisibly, without ever being detected? The answer, recorded in a classified document called Special Report No. 142, was yes.Forty-three monitoring stations around the Bay measured the result, and the Army concluded that every one of the city's 800,000 residents exposed to the cloud was inhaling five thousand or more particles every minute they breathed. The material was tracked as far as twenty-three miles inland. Three weeks later, an unprecedented outbreak began at Stanford's hospital, then located in San Francisco. Eleven patients developed infections caused by Serratia marcescens, an organism so rarely seen in clinical practice that the doctors published the cases in the Archives of Internal Medicine, where they remain the first documented hospital outbreak of that bacterium anywhere in the world. One patient died: Edward J. Nevin, seventy-five years old, an Irish immigrant who had kept his naturalization papers framed on the wall.This episode traces the whole arc, from the doctrine of the "harmless" simulant and the Pentagon test that came before San Francisco, through the twenty-six years of silence, the 1977 Senate hearings that revealed 239 open-air tests over American soil, and the federal lawsuit brought by Nevin's grandson that ended with a court ruling the question was not one an American court had jurisdiction to decide. It widens to the zinc cadmium sulfide dispersals, the New York subway experiment, the crates sent to Norfolk, and the radiation and syphilis studies that shared the same logic — and it stays honest about what was never proven and now never can be, because the hospital isolates were thrown away before anyone knew they mattered.Email BrianJoin Our FREE NewsletterGet Brian's Books Leave Us A VoicemailVisit Our WebsiteHave a forgotten historical mystery, disturbing event, unsolved crime, or hidden conspiracy you think deserves investigation?Send your suggestions to [e-mail] History is a dark history podcast exploring unsolved mysteries, secret societies, historical conspiracies, lost civilizations, and the shadowy stories buried beneath the surface of the past.Follow the show and enable automatic downloads so you never miss a deep dive into history’s most unsettling secrets.Because sometimes the truth is darker than fiction.

Some stories were never meant to be told. Others were buried on purpose. This podcast digs them all up. Disturbing history peels back the layers of the past to uncover the strange, the sinister, and the stories that were never supposed to survive. From shadowy presidential secrets to government experiments that sound more like fiction than fact, this is history they hoped you'd forget. I'm Brian, investigator, author, and your guide through the dark corner.

of our collective memory. Each week I'll narrate some of the most chilling and little-known tales from history that will make you question everything you thought you knew. And here's the twist. Sometimes the history is disturbing to us. And sometimes, we have to disturb history itself,

just to get to the truth. If you like your facts with the side of fear, if you're not afraid to pull at threads, others leave alone. You're in the right place. History isn't just written by the victors. victors. Sometimes it's rewritten by the disturbed. In September of 1950, a Navy ship spent eight days off the coast of San Francisco, spraying live bacteria at the city through hoses mounted

on its deck. 43 monitoring stations were set up on shore to track where the cloud went. When the numbers came back, the Army's own assessment was that the release had worked, and the finding in the report was that every one of the 800,000 people in San Francisco who'd been under that cloud, breathing at a normal rate of 10 liters a minute, had inhaled 5,000 or more particles. That got written up as good news. Three weeks later, in a hospital in that city, a lab technician looked at a culture plate from a patient

and found something growing on it that was bright red. The men running the test weren't measuring how far a cloud drifts. They were measuring how many living organisms they'd managed to put inside the lungs of a specific number of American civilians. and the answer they came back with was essentially all of them, 5,000 apiece, minimum. None of what I'm about to tell you is a conspiracy theory. There's no hidden document you have to take on faith, no anonymous whistleblower, no chain of inference holding the thing together.

The United States Army disclosed this program itself, under oath, in front of a Senate subcommittee in 1977. The reports exist, the court records exist, and the medical journal article written by the doctors who watched the aftermath exists too, published in 1951 in the American Medical Association's archives of internal medicine, and it's still cited in the infectious disease literature today because it happens to be the first recorded hospital outbreak of an organism that now makes people sick all over the world. Everything hard about this story lives in the gap between those last two facts.

To understand how a Navy ship ended up spraying back, bacteria at an American city. You have to understand what the men giving the orders believed was about to happen to that city. It's easy, 75 years out, to read early Cold War planning documents and see nothing but paranoia. They do read that way. But paranoia describes a mismatch between fear and reality, and in 1950, that mismatch wasn't as wide as hindsight makes it look. The people writing those documents had just watched the entire architecture of what was possible collapse and rebuild itself twice inside of a decade. Think about what an American defense planner had personally lived through by September of 1950. He'd watched a war in which cities were

deliberately burned down to their foundations from the air, Hamburg and Dresden and Tokyo, and that was accepted strategy. He'd watched two more cities erased by single aircraft carrying single weapons. He'd learned in the years after that the Japanese army had run a biological warfare enterprise in occupied Manchuria under the designation Unit 731, which experimented on prisoners, dissected them while they were alive, and dropped plague-infected fleas over Chinese towns. He'd also learned that his own government, instead of prosecuting the men responsible, had quietly traded them immunity for their research data. Then in August of 1949, four years earlier than most American estimates had predicted, the Soviet Union detonated an atomic bomb. And in June of

1950 the North Korean Army came across the parallel with Soviet tanks and the Cold War stopped being a metaphor. So a very specific question came out of that atmosphere and honestly it's a reasonable one. If an enemy wanted to attack the United States without an air force capable of reaching it, how would he do it? The answer that emerged inside the American defense establishment was that he'd do it quietly. He wouldn't need bombers. He'd need a fishing trawler, a merchant vessel, a suitcase, a man with a spray canister and a bus ticket. Biological agents are the weapons of the materially weaker party, because a gram of the right organism,

correctly packaged in aerosol form, can in principle do what a warehouse full of high explosive can't, and it can be delivered by anybody who can rent a boat. The man who made that argument most influentially was Ira Baldwin, a bacteriologist out of the University of Wisconsin, who'd served as the first scientific director of the American, American Biological Warfare Program at Camp Dietrich in Maryland during the Second World War. In the late 40s, Baldwin argued that the United States was uniquely exposed to exactly this kind of attack.

An enormous coastline, open ports, no meaningful detection capability, and a population that had never once been asked to think about it. And he argued that you couldn't assess a vulnerability like that on paper. You had to test it. The Army Chemical Corps, which ran Dietrich, agreed. It set up a Special Operations Division to carry out those tests, which meant covert simulated attacks on real American targets, using organisms the Army judged to be harmless in order to find out what an actual enemy could actually accomplish.

The first target was the Pentagon. In August of 1949, Special Operations Division personnel walked into the largest office building in the world, found its air handling system and sprayed bacteria into it. The building's own ventilation carried the organisms through corridors and offices where tens of thousands of people worked. Sampling equipment measured how far it spread. The people who worked there were never told,

and the exercise proved exactly what it was designed to prove, which is that the nerve center of American military power could be seated by one guy with a briefcase. From there, the program scaled up. In April of 1950, five months before San Francisco, Navy ships released clouds of simulant off Norfolk, Virginia, testing the same idea at sea. That worked too.

And the logic of escalation is remorseless in any research program. You prove the method on a small target, then you ask what happens at scale, and the only way to answer that question is to go find yourself a bigger target. From a planner's point of view, San Francisco was close to ideal. It was a major seaport, which made it a plausible enemy objective in its own right, the embarkation point for the Pacific and the gateway for the men and materials headed to Korea. It sat on a peninsula with a reliable onshore wind,

so a cloud released offshore would go where you predicted it would go. Its topography was complicated enough, with hills and valleys and a dense urban core and an open western flank, that the data would actually teach you something instead of just confirming the obvious. and then there was the fog, which offered an advantage I find harder to get past than any of the others. An aerosol released into that fog would be invisible, indistinguishable from ordinary weather,

exactly the way it would look if a real adversary attacked. The concealment wasn't a side effect. It was part of the experimental design. There's one more piece of context here that people tend to skip past, and I'd argue it's the most important one. Four years earlier at Nuremberg, American judges had sat in judgment of German physicians who'd performed experiments on human beings without consent. The judgment those American judges wrote, delivered in August of 1947, opened with a set

of principles that's been known ever since as the Nuremberg Code, and its first line isn't ambiguous. The voluntary consent of the human subject is absolutely essential. Three years and one month after that, a Navy ship sprayed bacteria and, and, you know, at 800,000 Americans who'd never been asked anything. Now, the men who ordered sea spray didn't consider themselves to be violating the Nuremberg Code. And it's worth being precise about why, because the reason tells you something about how atrocities and near atrocities actually get built.

They didn't think of San Franciscans as human subjects at all. In their scheme, a human experiment was something you did to a person. You administered something to a body. You observed the body. you recorded the result. Sea spray, the way they understood it, was an engineering test of dispersal characteristics. The subject of the experiment was the aerosol.

The city was terrain. And the people were part of the terrain, the same way the hills and the buildings and the wind were part of the terrain. That whole reframing rested on one load-bearing assumption and everything that follows in this story. The 11 patients and the dead man and the lawsuit and the 26 years of silence, depends on whether that assumption was true. The assumption was that the bacteria couldn't hurt anybody.

Every open-air biological test of the Cold War era depended on an idea that sounds almost absurd when you say it out loud, which is the idea of a safe germ. The technical term is simulant. A simulant is an organism you choose not for what it does to people, but for how it behaves in air, and the requirements are strict. It has to be roughly the same size and shape as a weaponized pathogen, so it drifts and settles and gets down into the lung the same way. It has to survive aerosolization and sunlight and drying out long enough to be measured. It has to be cheap to grow in enormous quantity. It has to be easy to detect and count once it lands. And it has to be,

in the judgment of the people releasing it, incapable of making anybody sick. For the American program, two organisms fit those criteria better than anything else, and both of them were aboard that ship. The first was bacillus globagi, a spore-forming soil bacterium closely related to bacillus anthrasis, which is anthrax. That relationship was the entire point. If you want to know how anthrax spores would move through a city, you release something that's physically almost identical to anthrax spores, but that doesn't, as far as anybody knew, cause anthrax.

Basilis Globiji is durable to an almost supernatural degree, and its spores survive for decades. Over the next 20 years, the Army released it at sites all over the country, in San Francisco and New York and Washington, and along the Pennsylvania Turnpike, among others. Modern taxonomy calls it Basilis Atrophelius. it's generally regarded as harmless to healthy people, though it's now understood like most environmental bacteria to be capable of causing disease in somebody whose immune system is badly compromised. The second organism is the one that gives this story its color, and I mean that literally.

Saratia Marsescence is a rod-shaped bacterium that lives in soil and water, on plants, in the damp corners of any building. It has one extraordinary property, which is that a lot of strains manufacture a brilliant blood-red pigment called prodigiosin. Let it grow on bread or a communion wafer or a bowl of polenta in a warm room, and it produces vivid crimson blotches that look uncannily like fresh blood. And for most of recorded European history, that's exactly what people thought they were looking at. The organism is a plausible explanation for a whole genre of medieval bleeding host miracles,

sacramental bread that appeared to weep blood, sometimes with lethal consequences for Jewish communities accused of desecrating it. In 1819, a red discoloration and a peasant's cornmeal in the Venado caused enough alarm that a young pharmacist named Bartolomeo Biceo went and investigated, identified the microbe responsible, and named it. Saraccia, after an Italian steamboat pioneer and Marquescence, for the way the red mass decayed. By the 20th century, that pigment had turned Saraccia

into one of the most useful organisms in science. If you want to trace where something goes, you use something you can see. Researchers used serratia marcessans as a biological dye. They swabbed it into volunteers' mouths and had them talk to trace how oral bacteria spread in speech. They introduced it before dental extractions to demonstrate how pulling a tooth-seeds bacteria into the bloodstream.

They released it into ventilation systems to map airflow. In 1906, something very like it was used to study the air inside the British House of Commons. By consensus, it was the workhorse harmless tracer of medical microbiology, and the professional literature said so for decades. That consensus was wrong. And here's what I keep coming back to. It was already visibly fraying when sea spray got authorized. The signals had been accumulating for years, scattered through the journals in ones and twos.

A 1943 paper described Saradia Marquesans recovered from a case of meningitis. French researchers in the 30s had reported the organism was pathogenic in animals and produced an endotoxin. In December of 1948 in Jerusalem, physicians described a 61-year-old man admitted with bronchid pneumonia who developed a urinary tract infection. And from that infection, a red-pigmented organism was repeatedly isolated in pure culture, very probably the first serracia urinary tract infection in the medical record. And there was, sitting in the American program's own files, a report of illness in a man who'd inhaled serratia.

None of that amounted to proof the organism was dangerous. It amounted to something I'd call more troubling than proof, which is an accumulation of anomalies that nobody had any incentive to organize into a conclusion. Stay tuned for more disturbing history. We'll be back after these messages. There's a familiar shape to institutional knowledge here. When a belief is load-bearing, when an entire testing program becomes impossible if the belief turns out to be false, the burden of proof quietly flips over.

The question stops being, have we established that this is safe, and becomes, has anybody proven it's dangerous? And in a field where nobody's collecting evidence, absence of evidence converts very smoothly into evidence of absence. There was a second gap in the reasoning too, and the trial 30 years later would eventually drag it into the open. The safety judgment came mostly from what happened to healthy adults exposed to the organism. Fort Dietrich had exposure data on volunteers, and the reported effects were mild and temporary. Some coughing, reddened eyes, a fever that passed in a few days. But a city isn't a room full of healthy adults. A city is also its newborns and its diabetics.

its cancer patients and its post-surgical wards. It's old men with failing hearts and catheters in their bladders. An organism a healthy immune system dispatches without even noticing is an entirely different proposition when you deliver it, in the millions, to a population that includes tens of thousands of people whose defenses are already compromised. And you're delivering it into the buildings where those people are concentrated because a hospital breeds the same air as the street.

The decision to go ahead anyway didn't come from some rogue laboratory. It came from the top. The officer with final authority over the Chemical Corps program was Major General Anthony McAuliffe, a genuine American hero, the man who at Bastogne in 1944, when encircling German forces asked him to surrender, famously answered with one word. Nuts.

McCallif personally approved San Francisco as the site. He personally approved San Francisco as the site. approved the selection of Saradia Marsesans after he and a fellow officer independently evaluated the strain safety. He wasn't bound by his technical advisors, and he knew it, and a federal appeals court would later note that he could have withheld approval for any reason at all, up to and including his own gut feeling that the thing was a bad idea. And then there's the detail I can't shake. McCalliffe did exercise that judgment, just in a different case. He rejected a proposal to run a test with the same organism

in the subways of New York City. He said no to New York and yes to San Francisco. We don't know why, and I'm not going to guess at a motive I can't document, but the fact establishes something beyond argument, which is that this wasn't a program where nobody thought about risk. Somebody at the top was weighing these tests against each other, deciding some were acceptable and some weren't. And he reached a considered conclusion

that spraying 800,000 unwitting Americans from a boat landed on the acceptable side of that line. In 1951, the year after San Francisco, wooden crates were deliberately contaminated with three organisms. Saracha marcessants, bacillus globagi, and Aspergillus Fumagodas,

which is a mold and a genuine, well-recognized cause of lethal infection in human beings. The crates were shipped to the Norfolk Naval Supply Center in Virginia, and the workers who unpacked them were the experiment. The rationale for including the, Aspergillis is preserved in the program's own documentation. The workforce at that supply center was predominantly African American. The researchers were operating on a belief drawn from the observed susceptibility of black Americans to coccidomycosis, that black workers would be more

vulnerable to fungal infection generally. And the report frames the whole exercise around the possibility that an enemy might deliberately target black laborers in the naval supply system in order to cripple it, using a more virulent fungus than the one being simulated. So a theory about racial vulnerability to disease gets used to justify exposing the very people, alleged to be most vulnerable without their knowledge, on the grounds that an enemy might someday do the same thing on purpose. Sea spray wasn't an aberration inside this program. It was the program working the way it was designed to work.

The operation ran from September 20th to set up. September 27, 1950. Eight days, six separate releases. The mechanics were unglamorous. A Navy vessel took up a position roughly two miles off the San Francisco coastline and ran a track parallel to the shore in the vicinity of the Golden Gate. A board equipment fed a bacterial slurry through large hoses fitted with atomizing nozzles, which broke the liquid into a mist of droplets, small enough to stay suspended for hours and to slip past the upper upper respiratory tract and large enough to carry a viable organism down into the lung. Each release ran about half an hour and produced a cloud that could stretch two miles

along the water before the wind took it inland. Everything about the timing was deliberate. Releases were coordinated with meteorological observation, wind speed and wind direction, and temperature and humidity, and the state of the fog. Conditions were monitored continuously and operations got a just or paused when the weather didn't favor transport toward the city. That crew wasn't spraying at random. They were waiting for the wind to be right, the way a bomber crew waits for a clear night,

and then they were releasing into it. On shore, 43 sampling stations were positioned across San Francisco and the surrounding area. The instrumentation was straightforward, devices that drew measured volumes of air across a collecting surface, from which you could culture and count the viable organisms. Saratia made that delightfully easy for the analysts, because a colony that grows out scarlet on a plate can't be mistaken for anything else. That's the whole reason it got picked.

The bacterium announced itself. From those counts, the analysts could reconstruct the cloud, where it went and how it thinned out, and how long it lingered and how the hills and the built environment channeled it. Detection ran well past the city limits. The material was tracked as far as 23 miles from the release. over the bay and into East Bay communities that had no relationship to the test at all, except that they happen to be downwind of it. If you know the airborne concentration at a given place in time, and you know the volume of air a resting adult moves through his lungs in a minute,

you can calculate an inhaled dose. That calculation is the deliverable. The report used the standard figure of 10 liters a minute for normal breathing and concluded that every one of the 800,000 people under that cloud was inhaling 5,000 or more particles a minute. To grasp what that means operationally, hold it up against the numbers that show up in modern risk assessments for inhalational anthrax.

The doses discussed in that literature, the quantities of spores associated with real infection risk in a human lung, are of the same general order of magnitude, which is to say that if the ship had been loaded with bacillus anthrasis, instead of its harmless-looking cousin, what got delivered that week wouldn't have been a near miss. It would have been an attack and an effective one, carried out by a handful of sailors with hoses,

and it would have been unattributable because it would have looked exactly like fog. That was the finding, and that was the point. Sea spray demonstrated with data that a major American city could be blanketed from a small boat offshore, that nobody would notice, and that essentially the entire population could be dosed in a single week. Read as an experiment, it was a complete success.

Red as anything else, it's a document in which the exposure of hundreds of thousands of people appears exclusively as a performance metric. Now, here's what's in the record and what isn't. Special Report No. 142 recorded the time of each release, the temperature, the wind speed, the humidity, the concentrations at each of the 43 stations, and the calculated inhaled dose. It recorded no provision for medical surveillance of the exposed population. There was no arrangement with San Francisco hospitals to watch for unusual infections. There was no notification to city or state health authorities.

There was no follow-up of any kind. Nobody planned to look. That's not an oversight in the ordinary sense, because it follows directly from the framing. If the organisms are harmless, there's nothing to watch for. And building a surveillance apparatus does nothing except create a problem. paper trail and risk exposing a classified operation. Leaving the monitoring out wasn't carelessness on anybody's part. It was consistency. It also had a consequence that's echoed for 75 years now, and no court has ever been able to resolve it, because nobody was watching.

Nobody knows what happened. Except in one building, where, three weeks later, somebody happened to look. Stanford's Teaching Hospital was in San Francisco in 1950. The medical school hadn't moved down the peninsula to Palo Alto yet. It sat in the city, and it took the city's patients, and it was a good hospital with a good laboratory, and crucially for this story, staff who were curious. Sometime that autumn, a technician looked at a culture plate from a patient's urine and saw something that shouldn't have been there. The organism was a gram-negative rod, and it was growing scarlet.

Physicians in 1950 didn't go hunting for Saratia Marquesans in a urine specimen, because serratia marcessants didn't cause urinary tract infections. That was the textbook position. The organism was the harmless red tracer, the thing you used to study infection, not the thing that caused it. Finding it in a patient was closer to a laboratory curiosity than a diagnosis. Finding it in patient after patient was something else. The index case was a man named Edward J. Nevin.

He was 75, an Irish immigrant who came to him. kept his naturalization papers framed on the living room wall. He had arteriosclerotic heart disease. About a month before the Navy started spraying, he developed acute urinary retention, and a catheter had been placed to relieve it. Roughly a month after that, in the general vicinity of the tests,

he went in for surgery to remove his prostate. Afterward, he developed a urinary tract infection, and the lab recovered a red pigmented gram-negative rod from his urine. The organism didn't stay in his urinary tract. It got into his bloodstream, and the same isolate came back from blood cultures. From the blood, it seated a heart valve, and he developed endocarditis, an infection of the lining of the heart, which in 1950, with the antibiotics then available, and against an organism nobody understood, was very close to a death sentence.

He died after spending 21 days in the hospital. He wasn't the only one. Around that same September and October window, four more red pigmented isolates came out of the urine of other patients. Between November of 1950 and February of 1951, six more. Eleven cases total in patients ranging from 29 to 78 years old. Ten of them survived, some after long and miserable hospitalizations. One didn't. The hospital's response was, by the standards of the era, exemplary, something genuine,

unprecedented was happening in their wards and they went after it. Physicians tried to trace the source and couldn't find one, and in 1951 of them, Richard P. Wheat, Aaron Zuckerman, and Lowell A. Rance published the series in the Archives of Internal Medicine under the title, Infection, due to chromobacteria, report of 11 cases. That paper is a landmark, and not for anything having to do with the Navy. It's the first described hospital outbreak of Saratia Marquezzi. in the medical literature. Every subsequent investigation of Sarasha as a healthcare-associated pathogen,

and there have been hundreds of them, from neonatal intensive care units to cardiac surgery wards to contaminated syringes and soap dispensers and nebulizers traces its citation lineage back to Wheat, Zuckerman, and Rance. 1951.

Page 461. The doctors who wrote it had no idea that three weeks before their first patient's culture, came back red, the United States Navy had sprayed billions of that exact species over their city and over their hospital and into the air their patients were breathing. They wouldn't find out for a quarter of a century. So did Operation C spray kill Edward Nevin? The honest answer, the one that survived a federal trial, an appeal, 75 years of scholarship, and the entire subsequent development of medical microbiology, is that nobody knows. That's unsatisfying.

It's also the only defensible answer. The case for a connection is largely a case about coincidence, and it isn't a weak one. An organism vanishingly rare in clinical practice showed up in 11 patients at one San Francisco hospital, starting within weeks of an operation that had deliberately saturated San Francisco with that organism and calculated that it deposited thousands of particles in the lungs of essentially every resident. The 1951 paper documents the first such hour, outbreak anywhere in the world. The hospital investigated and couldn't identify an internal source.

Cases of pneumonia in San Francisco rose in the period after the release. And in the decades afterward, physicians in the Bay Area would wonder out loud whether the region had inherited an altered microbial landscape, whether the heart valve infections and the severe infections seen among intravenous drug users in the 60s and 70s, and outbreaks as recent as one traced in 2001 to a Walnut Creek compounding pharmacy that killed three people might all trace back to descendants of what the Navy released. The San Francisco Chronicle reported that speculation in 2004. It's still speculation.

It's never been demonstrated. The case against a connection is also strong, and it was the government's case at trial. All 11 patients had undergone urinary tract instrumentation, catheters or cystoscopy or surgery, And that's the classic profile of a hospital-acquired infection, where the organism gets introduced directly past the body's defenses on a piece of equipment rather than inhaled out of the air. Stay tuned for more disturbing history. We'll be back after these messages. No other hospital in San Francisco reported a comparable cluster.

And if a citywide aerosol were seeding infections, you'd expect a citywide signal. Edward Nevin's own catheterization came before the spraying. Sarasha is a ubiquitous environmental organism that lives happily in the damp of any building. And the 50s and 60s saw these infections rise steeply across the entire United States and Europe. In hospitals with no connection to any test whatsoever, for reasons the infection control literature has since explained thoroughly, including the explosion of invasive devices, the widespread use of antibiotics that cleared the field of competitors, and the concentration of critically ill patients in intensive care.

And then there's the fact that ends the argument. The strains isolated from those 11 patients at Stanford were never archived. They were thrown out, the way clinical isolates routinely were. They were never compared to anything, not in 1951 when nobody knew there was anything to compare them to. And not in 1981 when the family sued. Because by then, they hadn't existed for 30 years. So the one experiment that could settle this can never be run.

Today it'd be trivial. Sequence the patient isolates. Sequence the military strain. Read the answer off a screen in an afternoon. The technology showed up four decades too late for material that had already gone into an autoclave. I'd call that its own category of harm, separate from whatever the bacteria did or didn't do. Secrecy didn't just conceal a fact from the public. It destroyed the ability to ever establish the fact, and not through any deliberate cover-up of evidence,

but through the ordinary operation of a system in which the people who could have preserved the evidence had no idea it was evidence. The doctors at Stanford would have kept those isolates under lock and key if anybody had told them what happened offshore. Nobody told them. So the isolates went in the bin, the question closed permanently, and every discussion of Operation C spray since, including this one, has had to be conducted in the subjunctive. The Nevin family would spend years and a lot of money learning that lesson in a federal court room. But first somebody had to tell them.

For 26 years, the only people on Earth who knew San Francisco had been sprayed were the people who'd sprayed it. And the program didn't stop. Between 1949 and 1969, the Army conducted 239 open-air tests of biological agents over American territory. 80 of those used the live bacteria, organisms researchers of the day believed to be harmless. The rest used inert chemical simulants standing in for microbes. Sea spray sat near the beginning of a 20-year experimental campaign, whose test sites were American cities, American highways, American transit systems, and American countryside. What ended it wasn't conscience.

It was a change in the political weather so complete that the old arrangements became structurally impossible to make. maintain. President Richard Nixon renounced offensive biological weapons in November of 1969 and ordered the stockpile destroyed. That destruction was finished by 1973 and the Biological Weapons Convention opened for signature in 1972 and entered into force in 1975. That removed the mission, but it didn't touch the secrets. What went after the secrets was Watergate and everything that came Moffett, the collapse of the presumption that national security agencies would police themselves, the Church Committee's investigation of the intelligence services, and the sudden unprecedented willingness of Congress to demand files. Reporters started pulling threads. In 1975,

the Rockefeller Commission and Senate investigators exposed M.K. Ultra, and with it the death of Frank Olson, a scientist at Fort Dietrich, who'd been dosed with LSD without his knowledge in 1953, and who went out of a New York hotel window nine days later. Once the Dietrich door came open, it didn't close again. In 1975, a Senate subcommittee questioned a Fort Dietrich physicist named Charles Censign about the New York subway experiment. And when Senator Gary Hart asked him whether the city's officials or its passengers had known, Cincinnati said he didn't believe so.

Through 1976 and into early 1977, the press reported that the Army had to the Army had run germ warfare tests over American cities. On the 22nd of December, 1976, a trial lawyer in his early 30s named Edward Nevin III was riding the commuter train into the city reading the San Francisco Chronicle. And he turned to the back page and found in an article about secret military testing,

his own grandfather named as the only person who died. He'd been nine years old in 1950, waiting in the family's 1941 Chevy outside the hospital while the adults went in. In February of 1977, the Army delivered an accounting of its open-air testing program to Congress. And later that year, the Senate Subcommittee on Health and Scientific Research,

under the Committee on Human Resources, in hearings devoted to biological testing involving human subjects by the Department of Defense, put the whole thing on the public record. What came out of those hearings wasn't a single scandal. It was an Atlas. San Francisco in 1950. Saradia dispersed over Panama City and Key West Florida through the 50s.

Basilis Globiji released in San Francisco, New York, Washington, and along the Pennsylvania turnpike. Zinc cadmium sulfide, fluorescent particles containing a compound of a heavy metal, sprayed over Minneapolis, St. Louis, Corpus Christi, Fort Wayne, and dozens of other locations, and dispersed from aircraft across an area running from the Rockies to the Atlantic, and from Canada to the Gulf of Mexico, with particles detected more than a thousand miles from where they were released. In 1965, Basilis Globiji released at Washington National Airport

and at the city's Greyhound Bus Terminal, exposing more than 130 travelers who then carried the organism to 39 cities in seven states over the following two weeks. In 1966, light bulbs filled with Basilis Subtilis, variant Niger, dropped onto the tracks of Midtown Manhattan, subway stations. The bacteria pulled for miles through the tunnels by the trains themselves. The report on that one written in January of 1968 concluded in language that requires no interpretation from me that a similar covert attack with an actual disease-causing agent during

rush hour could be expected to expose large numbers of people to infection and subsequent illness or death. On San Francisco specifically, Army witnesses acknowledged the pneumonia increase in the city after the release and testified that any link to the tests was entirely coincidental. They pointed to the arguments I've already laid out. No other hospital reported a comparable cluster, and every one of the 11 patients had undergone a urinary tract procedure, which pointed to a source inside the hospital, rather than in the air above it. Those are legitimate arguments, and they may well be correct. But they come with an asterisk, the Army's own conduct, attached to them, and you can't remove it. The reason nobody could produce evidence of citywide harm

is that no mechanism ever existed to detect citywide harm. The institution that chose not to look was now offering the absence of findings as reassurance. For the people who'd lived through it, none of this was abstract. Somewhere in San Francisco in 1977, members of a family read a newspaper account of germ warfare experiments over American cities, and a date attached itself to a memory. Their patriarch had gone into the hospital in the autumn of 1950. He died there of an infection nobody could explain, caused by a red bacterium the doctors had barely heard of, and it left his widow ruined by the medical bills. His grandson, named after him, was a lawyer. Edward J. Nevin III filed suit in the United States District Court for the North

Northern District of California, representing 67 members of his own family, the children and grandchildren and great-grandchildren of the man who died in 1950. The action was brought under the Federal TOR Claims Act, the statute by which Congress waived the government's sovereign immunity and allowed citizens to sue the United States for negligence. The claim was that Edward Nevin had died because his government sprayed him. Nevin was clear-eyed from the start about what he was doing. Several weeks into the trial, he cut the damages sought from $11 million down to $300,000, and said publicly that the family was concerned with principal rather than money, and that what motivated them was the prospect of a government running potentially fatal tests

on an uninformed public. He said later, more bluntly, that he'd known all along it wasn't a case you could win against the United States government, but that it was a story that needed to be told. And because the federal tort claims act provides for no jury, the jury box in Judge Samuel Conte's courtroom sat empty and the national press sat in it instead. That was the point. The family filled the gallery most days of the trial. They called the dead man their patriarch. They testified that they'd never questioned the cause of his death, that it had simply been one of those things that happens to an old man after surgery until the newspaper articles appeared. The government's defense stood on three pillars, and none of them was frivolous. The first was

scientific. The Army had believed in 1950 that Saratia Marcus Sins wasn't a human pathogen, and that belief was consistent with the scientific consensus of the time. The government's witnesses, including a physician from the biological warfare unit at Fort Detrick, testified to the safety data, including tests in which volunteers were exposed to millions of organisms, and suffered nothing worse than temporary coughing, red eyes, and a fever that cleared in a few days. An assistant United States attorney reportedly held up a pen in court and told the judge that every atom in that pen could spontaneously rise and rotate, and that this was about as likely as the bacteria having killed anybody. The second was causal. Even if the organism could hurt someone, the plaintiffs

had to prove that this organism, from that ship, killed this man. And they could, for the reason I've already given. The isolates were gone. What was left was a coincidence in time and a hospital population with catheters in it. The third pillar was legal, and it was the one that mattered. The Federal Tort Claims Act contains an exception, the discretionary function exception, which withdraws the government's consent to be sued over decisions involving policy judgment. The theory behind it is constitutional. Courts shouldn't use tort law to second second-guessed the political and strategic choices of the executive branch, because a nation whose defense planning can be litigated by any injured citizen

is a nation whose defense planning gets made by judges. On the first day, Judge Conti interrupted to ask why the test couldn't have been carried out somewhere unpopulated. Why not an uninhabited island, he asked, or Coosbe, Oregon? The government's attorney, John Kern, gave him the honest answer, which is that the wind currents thrown off by tall buildings, can't be reproduced in the countryside, so results from a rural site wouldn't have been comparable, which means the city was chosen precisely because it was a city.

The purpose of sea spray was to learn what happens when you attack an urban population, and a stretch of empty coast can't teach you that. The people weren't incidental to where the test happened. They were the subject matter. The evidence also established something the government's own safety case had quietly stepped around, which is that the bacteria could harm a small, group of people whose natural immunities were compromised, which is to say exactly the people who are in hospitals. Testifying for the family was Matthew Messelson, the Harvard molecular

biologist who'd served as a government arms control consultant, and whose arguments to the Nixon administration helped end the American biological weapons program. Messelson and other scientists testified that the serratia found in Edward Nevin's blood was most likely the same serratia the military had sprayed over the city, and that the army should have recognized the organism's potential to cause disease. Messelson has also described in interviews given decades later how he first came to suspect a connection. Roughly 15 years after the test, a laboratory assistant told him that her boyfriend had worked at the Navy's biological laboratory facility in Oakland, and that the base commander had assembled the staff to tell them a recent test had probably killed a man,

and that anyone who talked about it publicly would never work again. again in the United States. And outside the courtroom, an Army General challenged the plaintiff's lawyer to a fistfight. By the lawyer's own account, that general wasn't a villain performing villainy. He was furious that anybody would question the patriotism of men who'd done what they did, believing they were protecting the country. He felt like a hero, and he couldn't understand why this lawyer kept calling him something else. That's the whole story in miniature.

There's no scene in the history of Operation C-spray, where some of the story, somebody cackles. The men aboard the ship were sailors following orders. The men who designed the test were scientists trying to find out how badly their country could be hurt. The general who approved it had once told a German army to go to hell in one syllable. Every individual in that chain believed he was defending civilians from a catastrophe. And the sum of all that sincerity was 800,000 people breathing something they never agreed to breathe. And 11 patients in a hospital ward and a widow with bills she couldn't pay. On May 20th, 1981, Judge Conte ruled for the government on every issue.

He found the plaintiffs hadn't established that the Army's bacteria caused Edward Nevin's death. He found, in part, that the government was immune from suit in any event, because the decisions at issue fell inside the discretionary function exception. Stay tuned for more disturbing history. We'll be back after these messages. General McCalliffe had exercised exactly the kind of policy judgment that exception protects. Weighing advice he was free to disregard and deciding as he decided differently for New York, that this test should go forward.

The Ninth Circuit heard the appeal and issued its opinion in January of 1983. It agreed the discretionary function exception applied, and then it took the reasoning one step further in a way that's legally elegant and for a family looking for a hearing, brutal. Because the exception is jurisdictional, the district court had never possessed the authority to reach the merits at all. So the appeals court vacated the judgment and sent the case back with instructions to dismiss it for lack of subject matter jurisdiction. One judge dissented in part, arguing the district court did have jurisdiction, but he would have affirmed anyway on the ground that the finding that the army's bacteria hadn't caused the death was not clearly erroneous. The Nevin family took it to the Supreme Court of the United States. which declined to disturb the result.

Strip the language off the legal outcome, and here's what's left. Whether or not the United States government killed Edward Nevin, that question isn't one an American court may consider, because the decision to spray a city with bacteria is a policy choice, and policy choices aren't the business of tort law. Reasonable people can defend that doctrine in the abstract. It's genuinely meant to protect democratic accountability by keeping strategic decisions in the political branches instead of the courts.

But applied here it produces a closed circuit. The tests were secret, so nobody could object to them beforehand. Because they were secret, no medical surveillance existed, so no evidence of harm got gathered. Because no evidence got gathered, causation couldn't be proved afterward. And even if it could have been proved, the decision was discretionary, so the courthouse door was shut anyway. At every stage, the secrecy that made the harm possible also made the harm unprovable. And then the law treated the unprovability as an answer.

Operation C-spray is the clearest single case, but it isn't really a case about San Francisco. It's a case about a method, and once you accept that a civilian population can serve as an instrument, geography turns into a variable. The zinc-cadmium sulfide dispersals ran through the 50s and 60s. For years, the Army Chemical Corps and its contractors, Stanford University among them, released clouds of fluorescent zinc-cadmium sulfide particles over inhabited areas to trace how aerosols travel. Minneapolis, St. Louis, Corpus Christi, Fort Wayne, and 29 other urban and rural locations in the United States and Canada were used. In St. Louis, the test ran in 1953 and then again from 1963 to 1965.

with material released off rooftops and from vehicles driving through neighborhoods. The particles weren't biological agents. They were simulants, chosen for the same reason as Saratia, because you could see them and count them. But zinc cadmium sulfide contains cadmium, and cadmium compounds are classified as probable human carcinogens. The largest of those programs was Operation LAC for large area coverage, and it was the biggest test the chemical core ever ran. Aircraft laid down particle trails across a test area stretching from the Rockies to the Atlantic and from Canada to the Gulf of Mexico,

and it established successfully that thousands of square miles of a country could be covered with a biological agent. When residents of those cities found out what had been done above their heads, in the early 1990s, the outcry was severe enough that Congress asked the National Research Council to investigate. The resulting 1997 report is a model of what independent review is supposed to look like. It reconstructed the releases, estimated doses, and concluded that it found no evidence exposure at those levels could have made anybody sick. The highest-modeled lifetime lung cancer risk it calculated, for a person maximally exposed in St. Louis during both testing periods, came out on the order of one in a million.

The report also documented substantial uncertainty in the underlying data, and the army itself had acknowledged that its own risk estimates should be read as conditional rather than absolute. On the specific question of whether the fluorescent particle tests poisoned American cities, the answer from the most rigorous review ever conducted is probably not. What that review couldn't do, and what no review can do is retroactively supply the consent that was never sought. One critic's summary of the program is the one that stuck with me, which is that the Army had been using the country as an experimental laboratory. Operation White Coat is the counter-example, from 1954 to 1973, some 2,300 United States Army personnel,

most of them Seventh-day Adventists, conscientious objectors whose faith forbade them to bear arms but not to serve, volunteered as human subjects in biodefense research at Fort Dietrich. They were briefed on the risks. They gave informed consent. they could withdraw. They were exposed to agents including coxiella burnadi, which causes Q fever and teleremia. And the research that came out of it contributed to vaccines and treatments that exist today. White coat matters because it demolishes the most common defense of everything else on that list, which is that these were the standards of the time, and nobody thought about consent in the 1950s. Somebody did. In the same program, in the same decade, at the same installation,

The American military ran a human research effort with a consent process and ethics board would recognize today. The men who sprayed San Francisco weren't prisoners of their era's moral imagination. Their own colleagues in the next building were demonstrating the alternative, widen the frame one more time, and C-spray stops looking like a story about biological warfare at all. It becomes one instance of a pattern that ran straight through mid-century American institutional life, in which the population of the United States got treated, repeatedly, across unrelated agencies,

as a surface you could test questions on. The pattern isn't a conspiracy. There was no coordinating body. The Chemical Corps didn't consult the Atomic Energy Commission, which didn't consult the Public Health Service. What they shared was something more ordinary and more durable than a plot, which is a common set of assumptions about who gets asked. Start with the radiation experiments.

beginning in 1945 in connection with the Manhattan Project. 18 people were injected with plutonium to study how the body metabolized it. They were hospital patients, and they weren't told what they were being given. The first designated HP12 in the records was a black man named Eb Kade, a cement worker admitted after a car accident in Tennessee. In the years that followed, the constellation of studies grew. Children at a state school in Massachusetts were fed radioactive, isotopes in their breakfast cereal as part of a science club.

Pregnant women were given radioactive iron. Prisoners' testicles were irradiated to establish sterilization thresholds. In December of 1949 at the Hanford site in Washington State, a deliberate release known as the Green Run put a large quantity of radioactive iodine into the air over-inhabited country, apparently to calibrate instruments for detecting Soviet plutonium production. An atmospheric release conducted, like sea spray, over people who weren't informed. When President Clinton established the Advisory Committee on Human Radiation Experiments in 1994, its investigators worked through hundreds of thousands of pages and identified thousands of studies.

In October of 1995, the president apologized to the subjects and their families on behalf of the United States. Then there's M.K. Ultra. The Central Intelligence Agency's program of research into behavioral modification, running from 1953 into the 60s, included administering LSD and other drugs to unwitting subjects. Among them, the patrons of safe houses operated in San Francisco and New York, under the operation name Midnight Climax, where prostitutes on the agency payroll dosed men who'd come in for something else while officers watched through one-way glass. Its connection to this story isn't thematic.

It's literal. Frank Olson, the Dietrich scientist who died in 1953 after being given LSD without his knowledge, worked in the same biological warfare establishment that ran sea spray. When director Richard Helms ordered the MK Ultra Files destroyed in 1973, the destruction was thorough enough that what we know today survives largely because a misfiled cash of financial records escaped the ship. shredder. Then there's Tuskegee. From 1932 to 1972, the United States Public Health Service followed 399 black men with untreated syphilis in Macon County, Alabama, along with a control group, in order to document the natural progression of the disease. The men were told

they were being treated for bad blood. They weren't treated at all. When penicillin became the standard cure in the 1940s, they weren't given it. And the studies administrators took steps to keep them from getting it anywhere else. That study ran for 40 years through the entire span of the biological testing program I've been describing. And it ended in 1972 only because a public health service employee named Peter Buxton, unable to get it stopped internally, went to the press, and Gene Heller of the Associated Press published it. President Clinton apologized in 1997 to the eight survivors who were still alive. And there's Guatemala. Between 9,000, 1946 and 1948, American researchers, some of them the same people who worked on Tuskegee,

deliberately infected Guatemalan prisoners, soldiers, psychiatric patients, and sex workers with syphilis and gonorrhea to test prophylaxis. A historian uncovered the program in 2010, and the United States apologized to Guatemala. Set all of that beside Operation C-spray and the shared logic gets easy to see. In every one of these cases, the subjects were selected because they were available, rather than because they were appropriate. Patients already in a hospital bed, children already in an institution, prisoners already confined, sharecroppers with no other doctor, a city that happened to sit downwind of open water. In every case, the researchers were pursuing an objective they sincerely believed was important, and often it genuinely was, because plutonium

metabolism data save the lives of nuclear workers and vulnerability assessment is a real and necessary discipline. In every case the harm was framed as speculative and the benefit is certain, which is exactly backwards from how you're supposed to weigh risk when the person carrying the risk hasn't agreed to carry it. And in every case, secrecy got justified by the importance of the work. And then that same secrecy prevented anybody from checking whether the work's premises were even true. The correction when it came came from disclosure rather than from conscience, from journalists, from a congressional subpoena, from an employee who wouldn't let it go. Tuskegee's exposure in 1972 produced the National Research Act of 1974 and the commission that wrote the Belmont

report in 1979, articulating respect for persons, beneficence, and justice as the foundations of human subjects research. Institutional review boards, the common rule, the whole modern architecture of informed consent, nearly all of it sits downstream of scandals like these. The rules exist because these things happened. That's not an inspiring lesson, but it's a load-bearing one. The protections that would have made Operation C spray impossible weren't the product of moral progress in the abstract. They were the product of somebody finding out. There's a final irony in this history. The central premise of Operation C-spray was that Saratia marcessants doesn't cause human disease.

By the mid-60s, that premise was dead. The literature accumulated case by case and outbreak by outbreak until the organism's status wasn't arguable anymore. Saratia marcessans is recognized today as a significant opportunistic human pathogen, a routine cause of urinary tract infection, pneumonia, bloodstream infection, meningitis, endocarditis, and wound and eye infection. It has an established affinity for exactly the environments where medicine gets practiced, colonizing soap dispensers, and disinfectant solutions and nebulizers, and pressure transducers,

and catheters, and pre-filled syringes. It has a particular reputation in neonatal intensive care units, where outbreaks have killed premature infants, and it carries multiple antibiotic resistance determinants and readily picks up more, which makes the infections it causes harder to treat now than they were in the 1970s. And the reference point for all of it, the citation at the head of the lineage, the first documented hospital outbreak of Saraccia Marcessans in the history of medicine, is Wheat, Zuckerman, and Rons, 1951. 11 patients at a San Francisco hospital in the weeks after the United States Navy

sprayed the city with the same organism. whatever the causal truth of that coincidence turns out to be, the verdict on the operation's reasoning is unambiguous. The safety judgment that authorized sea spray was wrong. It wasn't wrong in some subtle way that only modern methods could have revealed. It was wrong about the basic question. Can this organism make people sick? And the answer turned out to be yes.

And the evidence pointing toward yes was already scattered through the literature when the decision got made, and nobody assembled it, because assembling it would have made the program impossible. There's a lesson in that which outlives the Cold War, and it has nothing to do with germs. The men who approved a sea spray were never required to prove that Saratia was safe. They were required only to satisfy themselves that nobody had proven it dangerous. Those are radically different standards, and the gap between them is where almost every institutional catastrophe of the modern era has taken up residence. When a program's viability depends on a factual belief, the organization tends to defend the belief instead of testing it, and that's not corruption.

It's just the ordinary human preference for evidence that lets the work continue. Secrecy accelerates it enormously, because secrecy removes the outside critics who'd otherwise assemble the anomalies into a pattern. A classified program is a program where the only people qualified to notice the premise is wrong, are the people whose careers depend on it, being right. And then there's the word simulant, which is the piece of this that generalizes furthest. A simulant is defined by what it's for. It's a stand-in, a placeholder, a thing that isn't the real thing. That word does a quiet and enormous amount of work, because it converts the release of billions of living organisms into the human respiratory tract from an act into a procedure. The language of the program was full of those conversions. The city was a target area,

the exposure was a dosage. The 800,000 people were a population in the statistical sense, a denominator rather than a set of individuals with lungs. Edward Nevin, in the framework of Special Report No. 142, doesn't exist and couldn't exist, because there's no field in that document where a man like him could be recorded. That's how it's always done,

with vocabulary rather than with malice. The fog still comes in. It builds over the cold water and comes through the gate, spills over the headlands and climbs the hills, and the foghorns still call to each other across the bay. The city under it is bigger now, and richer, transformed almost past recognition from the working port that received six bacterial clouds in September of 1950. Nobody standing at Ocean Beach watching it arrive is thinking about a Navy vessel two miles out with hoses on its deck.

There's no memorial. There's never been an apology. No court and no commission has ever found as a matter of fact that Operation C spray harmed anybody, and none has ever found that it didn't. And there can never be either finding, because the material that would settle it got discarded as medical waste in 1951 by doctors who hadn't been told what they were holding. What's left is the record, and the record isn't in dispute.

Between September 20th and September 27th, 1950, The United States Navy released clouds of live bacteria from a ship off San Francisco. Six times, time to the wind. 43 monitoring stations recorded the result. The Army calculated that every one of the city's 800,000 exposed residents was inhaling 5,000 or more particles a minute. Nobody was told, before or after. It was one of 239 open-air tests conducted over American territory between 1949 and 1960s.

80 of which used the live bacteria. Within weeks of the release, an unprecedented outbreak of an organism the doctors had never seen began in a San Francisco hospital, killed a man named Edward Nevin and entered the medical literature as the first of its kind anywhere in the world. The connection was never proven and never disproven. The story surfaced in the press in December of 1976, 26 years later, and the Army laid the program out before the Senate the following year. And a federal court held in 1981 and an appeals court confirmed in 1983 that whether the United States had killed Edward Nevin wasn't a question an American court had jurisdiction to decide. The reason this one belongs in disturbing history isn't that

it's lurid. It's that it's boring. In the specific and terrible way, institutional decisions are boring. There's no cackling villain anywhere in it. There's a meteorological log. There's a table of particle counts. There's a general who'd already proved his physical courage to the entire world, weighing two proposals and approving one of them. There's a report with three stated objectives, not one of them malicious, and no section anywhere in that report where the 800,000 people breathing the aerosol appear as anything except the medium it moved through.

That's the disturbing part for me. Nobody ever had to decide to hurt a city. Somebody only had to decide the city was terrible. rain.