Moscow blamed bad meat, and a Harvard scientist believed them.

April 2, 1979.

Sverdlovsk sat on the eastern slope of the Ural Mountains, on the line where Europe stops and Asia begins. It was a city of factories. Tanks came out of it. Heavy machinery came out of it. During the Second World War, whole plants had been loaded onto trains in the west and rebuilt there, beyond the reach of German bombers, and many never went back. By the late 1970s more than a million people lived in Sverdlovsk, and almost none of them had ever met a foreigner. The city was closed. Visitors from the West were not permitted.

Americans knew the name for one reason. On May 1, 1960, a Soviet missile brought down a U-2 spy plane near Sverdlovsk, and its pilot, Francis Gary Powers, came down alive by parachute. The CIA had been flying over the Urals because there was a great deal there it wanted to see.

Some of it was in the southern part of the city, in a walled military district known as Compound 19. It belonged to the Ministry of Defense. The people who lived around it knew it was military and knew not to ask. Officially it worked on microbiology. A Soviet scientist who later ran part of the country's biological weapons program, Ken Alibek, would write after he defected to the United States that Compound 19 made anthrax, dried and milled into a powder fine enough to be breathed.

Seven years before, in 1972, the Soviet Union had signed the Biological Weapons Convention alongside the United States and Britain. The treaty banned developing, producing, or stockpiling biological weapons. President Nixon had already ended the American offensive program in 1969. The convention went into force in 1975. Compound 19 kept working.

Southeast of the compound, across the road, stood a ceramics factory. Its workers made tile and pipe, worked in shifts, and walked to the bus at the end of them. Beyond the factory ran apartment blocks, then the edge of the city, then villages with small herds of sheep and cattle, then open country.

Anthrax is a bacterium that lives in soil and forms spores. Spores survive for decades. Farmers and herders have known the disease for thousands of years, mostly as a skin infection or from eating meat from a sick animal. Inhaled, it is a different disease. The spores settle deep in the lungs, travel to the lymph nodes of the chest, and germinate there. For a day or two the victim feels like he has the flu. Then the bacteria release their toxin, and the body fails fast. Before antibiotics and intensive care, inhalation anthrax killed almost everyone it reached. Once symptoms became severe, treatment rarely helped.

The account of what happened inside Compound 19 that spring comes from Alibek, who said he heard it within the program. The drying equipment at the plant ran exhaust through filters, and the filters kept the spores inside the building. During one shift, a technician removed a clogged filter. He wrote a note saying it was out. According to Alibek, the supervisor did not enter it in the logbook. The next shift arrived, read the log, found nothing out of order, and started the machines.

Whatever the exact sequence was, something left the building. The wind that Monday was blowing out of the north and northwest, toward the southeast. It carried whatever came out of the compound across the road, over the ceramics factory, over the apartment blocks, and on into the countryside.

Nothing happened that anyone could see. There was no smoke, no smell, no sound.

Within days, men started getting sick. They had fevers, chills, aches, a cough. Some went to work anyway. Then they got worse, abruptly, with high fever and trouble breathing, and their families took them to clinics. Clinics sent them to hospitals. Some died within a day of admission. Some died at home before anyone thought to call for help. Many of them had worked at the ceramics factory. Others lived or worked along the same line running southeast from the compound wall.

The doctors noticed something odd about who was dying. They were nearly all adults, most of them men. There were no children. Doctors at first could not explain it.

The patients were gathered at City Hospital 40, which had an infectious disease section. A pathologist there named Faina Abramova performed the autopsies. She found blood in the membranes covering the brain, a pattern pathologists call the cardinal's cap. That was consistent with anthrax. What she found in the chest mattered more. The lymph nodes between the lungs were swollen and hemorrhagic. That damage meant the spores had come in through the lungs. Bad meat does not do that. Breathing does.

Abramova and a colleague, Lev Grinberg, kept detailed notes. They preserved tissue samples from the bodies.

Officials from Moscow arrived. The explanation they gave the city was that people had eaten anthrax-contaminated meat from private sellers. Residents were warned against buying meat outside state stores. Stray dogs were killed. Crews washed down buildings and trees in the affected district. Some dirt roads were paved over. Residents were offered vaccinations. Families of the dead were told their relatives had died of the gastrointestinal form of the disease, and the bodies were buried quickly, in a section of a city cemetery set aside for them, treated with chlorinated lime.

Security officers collected the hospital records. Abramova and Grinberg did not hand over everything. They kept their specimens and their notes, and they kept them hidden.

Cases kept turning up for weeks, into May. Then they stopped.

Out in the villages southeast of the city, sheep and cattle had died that spring in numbers the farmers remembered. The villages lay along the same line as the ceramics factory, dozens of kilometers out.

Word of the outbreak leaked to the West through an émigré publication later in 1979, and American intelligence picked it up. In 1980 the United States government raised it publicly and asked whether the Soviet Union had violated the convention. Moscow answered that there had been a natural outbreak caused by contaminated meat, and that there was nothing more to say.

In the United States, the most prominent outside voice on the question was Matthew Meselson, a Harvard molecular biologist. He was one of the scientists whose work had helped persuade the Nixon administration to give up biological weapons, and he knew the disease. He thought the meat explanation was possible. Gastrointestinal anthrax outbreaks were a real thing in rural parts of the world. He said the evidence did not settle the question.

In 1988, Soviet health officials came to the United States and presented their case in person to American scientists. They brought data. The data described contaminated meat, gastrointestinal disease, and a natural outbreak. Meselson found the presentation plausible.

The Soviet Union ended in December 1991. In 1979, the Communist Party chief of the Sverdlovsk region had been Boris Yeltsin. By 1992 he was president of Russia, and in a newspaper interview that year he said the KGB had acknowledged that the outbreak came from military work.

That June, Meselson went to Yekaterinburg, which had taken back its old prerevolutionary name. He brought a team. It included Jeanne Guillemin, an anthropologist at Boston College and his wife. Guillemin went door to door. She found families of the dead and asked where each victim had lived and where each had worked or been on the days in early April. The team went to the cemetery and read the headstones. They tracked down what remained of the medical files.

Abramova and Grinberg brought out the notes and the tissue samples they had hidden in 1979.

The team plotted every case on a map. Plotted by home address, the cases were scattered. Plotted by where each person had been during the day around April 2, they fell into a narrow zone running southeast from Compound 19, a few kilometers long. The animal deaths in the villages continued along the same axis. The team checked weather records for the first days of April. On April 2, the wind had blown along exactly that line.

A meat outbreak would have followed the meat, wherever it was sold and carried. This followed the wind. The children had been inside schools and homes that day. The men had been outside, at work, along the plume.

The pathology came out in 1993. The epidemiology came out in the journal Science in 1994. The study counted seventy-seven cases. Sixty-six of them died. The authors noted that the real numbers could be higher, because some records were never found.

The team concluded that the release had been airborne, that it came from the military facility, and that it had very likely happened on April 2, 1979. Meselson, who had spent years publicly saying the Soviet explanation could be true, put his name on the paper that ended it.

No one was charged. No Soviet or Russian official was publicly disciplined over the release. Russia passed legislation providing some compensation to families of the victims.

Ken Alibek left for the United States in 1992 and testified about the Soviet program before Congress. His book, Biohazard, was published in 1999.

Russia never released its official investigation of the outbreak. The hospital records taken in 1979 were never returned. Some Soviet officials who had defended the meat explanation kept defending it for years afterward.

Faina Abramova and Lev Grinberg's preserved tissue was later studied with genetic methods. The samples showed several strains of anthrax mixed together, which is not what a natural outbreak looks like.

Jeanne Guillemin published her account of the investigation in 1999. Meselson's work on the case is still cited as one of the cleanest pieces of outbreak forensics on record.

The victims are still in the cemetery section they were given, their headstones carrying dates from the spring of 1979.

Compound 19 is still a military installation, and no outside investigator has ever been allowed through its gate.