It was 11 minutes after noon on the third Friday in March, and Sherif Zaki was in a meeting at the Centers for Disease Control and Prevention (CDC) when he got a message on his pager. "I can’t believe it," the message said, "but it looks as though Tom’s group has isolated a coronavirus. The cells were ‘fried’ by the microwave, but I’m pretty certain (90 percent) that’s the result. Call me, I’m waiting on you to look before I pass on the info. Cynthia."

Zaki is chief of infectious disease pathology at the CDC. Pathology, the study of the nature and effects of disease, is not one of the heart-pounding specialties in medicine. It’s a field for brainy people who are happy to spend long periods of time at the microscope, scanning cellular landscapes for the unusual or telling feature. In their tolerance for high-risk monotony, they resemble air traffic controllers or lifeguards. As in those jobs, they are occasionally called on to move quickly. Zaki kept his counsel—the "Tom" of the message was sitting a few feet away—but messaged back: "I’ll be right down." He excused himself from the meeting.

Much of what can be said of pathologists also holds for electron microscopists. In 19 years at the CDC, Cynthia Goldsmith, author of the text message to Zaki, had looked at a lot of hostile territory and picked out a lot of bad actors at a magnification of 40,000x. She was among the first to take a picture of the Sin Nombre hantavirus responsible for cases of fatal pneumonia on a Navajo reservation in the Southwest in 1993. In 1999, she was first to identify the Nipah virus, which killed about a hundred pig farmers and slaughterhouse workers in Malaysia and Singapore. Both times, though, other lab tests had given her hints of what to look for. This was different. She was looking for the possible agent of severe acute respiratory syndrome (SARS), a contagious, sometimes fatal infection that had appeared on two continents half a planet apart. Nine days earlier, the World Health Organization (W.H.O.) had issued a "global alert" about the disease. She peered through the electron microscope at a virus originally taken from the throat of a SARS patient in Asia and grown in a flask of cells at the CDC. What she was seeing wasn’t what people said she should be seeing. Her heart raced as she and Zaki studied the images on a green phosphorescent screen.

 Coronaviruses—the name comes from the spikelike formations on the virus surface that sometimes resemble a corona, or crown—were far down any list of candidates for the cause of SARS. Coronaviruses can cause colds (though not most colds, which are caused by rhinoviruses) and, in premature infants, pneumonia. But in general, coronaviruses are so unthreatening to human health that the 2,629-page Harrison’s Principles of Internal Medicine, the world’s best-selling English-language medical textbook, devotes a mere six paragraphs to them.

Yet Goldsmith was certain that she was indeed looking at a coronavirus. In 15 minutes, Zaki was convinced too. He and Goldsmith went down the hall, where Charles Humphrey, another electron microscopist, was looking at a virus sample from the same patient, Carlo Urbani. The first W.H.O. physician to investigate a SARS case, at the Vietnam French Hospital of Hanoi, Urbani would die eight days after the CDC researchers made their observation. Humphrey used a negative-stain technique—basically a form of backlighting—to outline the material. The virus sample was in poor condition, which made identification difficult. Nevertheless, Zaki, with studied neutrality, asked Humphrey what he thought he was seeing. As Zaki later explained, "Part of science is to do things in blinded fashion. I didn’t want to ask him a leading question. I was trying to avoid that at all costs."

Humphrey has been looking at infectious agents with electron microscopes since 1968. "It could be an influenza [virus] or a coronavirus," he told Zaki. "I was not quite ready to lean one way or the other," he said later. "It had characteristics of both." After Zaki and Goldsmith peered at Humphrey’s images, they took him to look at Goldsmith’s. By the middle of the afternoon, the trio was ready to share its conclusion with CDC colleagues: it was a coronavirus. Three days later, the CDC told the world.

In retrospect, it’s the excitement of the discovery that medical researchers remember. But at the time, they also felt apprehension. A new, often fatal disease was loose in several densely populated cities in China, and among the more frequent victims were medical workers.

The cause of the disease is a virus around 100 nanometers in diameter, or four-millionths of an inch. Genetically, the SARS virus (SARS-CoV) doesn’t closely resemble any of the dozen well-studied coronaviruses known to infect animals or people. It doesn’t even fall into one of the genus’s three broad genetic groups, forming instead a new branch on the family tree. Its origin is unknown—and so, in a sense, is its destination. Biologists haven’t yet charted the full range of human tissue it can inhabit or attack. And nobody knows whether it’s here to stay as a permanent disease that human flesh is heir to.

What scientists do know about coronaviruses in general, and SARS-CoV in particular, suggests that infection may differ considerably from victim to victim, persist over time and be difficult to vaccinate against. SARS-CoV stores genetic information in single-stranded RNA, a less stable and more mutable molecule than the double-stranded DNA used by fungi, human beings and everything in between. Coronaviruses have a larger genome, or collection of hereditary material, than any RNA virus studied so far. In addition, they carry an unusual enzyme that permits two sister viruses to swap genes if they happen to find themselves infecting the same cell. That capacity to form "recombinants," or hybrids, as well as the virus’s large genome, enable the genus to easily gain or lose traits. Such traits may include the ability to infect new species, elude the immune system and change residence in the body over time.

The story of transmissible gastroenteritis virus in pigs demonstrates how coronaviruses acquire new powers. The disease, known since the 1940s, causes severe diarrhea in piglets. Periodic outbreaks have killed whole generations of animals on some farms. In 1989, farmers in Europe began noticing a new respiratory infection in pigs. The cause turned out to be a genetically altered form of the gastroenteritis virus that had evolved the capacity to invade the lungs. Coronaviruses are changelings, multitaskers, rule breakers. Bovine coronavirus causes several different diseases in cattle. In calves, it causes severe diarrhea; in yearlings, a pneumonia called shipping fever; in adult cows, a dysentery-like illness.

Coronaviruses are versatile in other ways too, with some strains able to infect more than one species. A study two years ago showed that a coronavirus isolated from cattle could also infect baby turkeys, though not, curiously, baby chickens. "Coronaviruses may be much more promiscuous than we originally thought," says Linda Saif, a veterinary scientist and virologist at Ohio State University.

Scientists have only begun to learn the rules of engagement the SARS coronavirus follows. Like many of its kin, it appears to be a lung-and-gut bug; people die from lung damage; about one-fifth of its victims also have vomiting and diarrhea. But SARS-CoV behaves unlike many respiratory viruses. For one thing, the disease it causes develops slowly. Also, there’s an almost miraculous sparing of children. In the recent SARS outbreak, few children became ill and none under age 16 died. Scientists don’t yet know why.

If SARS-CoV entered the human population from animals, it is by no means the first virus to make the jump between species. Measles, which has afflicted human beings for at least 2,000 years and still kills more than 700,000 people annually (mostly children), is caused by a virus whose closest relative causes rinderpest, a disease of cattle. The domestication of animals brought human beings and bovids together in large numbers, and some of the herd’s pathogens adapted to life in the herders. A similar leap ages ago may have introduced human populations to the smallpox virus, which has since been eradicated.

Perhaps the most important question about SARS—is it with us forever?—can’t yet be answered. According to preliminary reports, some exotic mammals in southern China that are caught and sold for food (including the masked palm civet) harbor a coronavirus identical to SARS-CoV with an important exception: the animal virus’s RNA has an additional 29 nucleotides, or chemical subunits. The similarity suggests that the SARS virus arose from the animal virus. If those 29 missing nucleotides hold the key to the emergence of SARS-CoV, its future may depend on how frequently that particular genetic deletion occurs. It may not happen again for decades, or centuries. Or it could happen next year. But even if the virus’s genetic material changes frequently, future epidemics may possibly be prevented merely by keeping people away from palm civets and other infected species.

Alternatively, SARS may behave like Ebola hemorrhagic fever, which appears periodically. Ebola emerged in 1976 in simultaneous outbreaks in Zaire and Sudan. The virus strikes in Africa every few years, killing 50 to 90 percent of the people it infects, and then vanishes. Despite great effort, scientists still haven’t found the natural animal host or reservoir for Ebola virus, and that makes it harder to prevent periodic outbreaks.

By early July, the W.H.O. declared that the outbreak was over. At last count, 8,399 people in 30 nations had been identified as "probable" SARS cases and 813 of them had died.

Of course, even though the SARS epidemic is officially over, the virus may actually still be with us. A few survivors are known to have carried it for months and may be contagious. It’s also conceivable that a handful of people with the disease have escaped detection. For those reasons, some medical experts believe that only a vaccine can rid humanity of SARS for certain. Making and testing one will require at least three years of work, says Gary Nabel, director of the vaccine research center at the National Institute of Allergy and Infectious Diseases. (The same is likely to be true for anti-SARS drugs.) Even so, animal coronavirus vaccines have a spotty record. Some provide only transient protection. Others, like the vaccine against feline coronavirus, can even worsen an infection under some circumstances. Until good drugs and an effective vaccine are available, the best approach to preventing the global spread of the disease is decidedly old-fashioned: identifying infected persons, isolating them until they recover and quarantining people who’ve had close contact with the victims. Those measures, applied assiduously in recent months and in many nations, appear to have accomplished something nearly unheard of in the history of medicine—halting an epidemic respiratory infection, at least temporarily.

For his part, the CDC’s Zaki is betting on SARS’s return. "I don’t see any reason why it shouldn’t come back," he says. "We can learn from history. If it happened once, it can happen again."

The flip side of such fatalism—or is it realism?—is that despite some predictions that the emergence of SARS augurs a new millennium of ever-accumulating human scourges, nothing about it is foreordained. We shouldn’t forget that thanks to sanitation, affluence and medicine, in many parts of the world far more infectious diseases have retreated than have emerged in the past century. The appearance of SARS, like so many important historical events, was the product of dozens, or hundreds, of small occurrences, many of them chance. It was neither inevitable nor entirely unexpected. It’s just what happened.