I had placed a video call to Relman, on Sunday afternoon, because I had hoped he might help me characterize the evidence for each theory. He said he saw several points in favor of zoonotic spillover. The first was that this was usually how new viruses emerged in people, and the literature suggested that animal crossovers are “happening far more than we know.” At the margins of human civilization, where villages pressed up against the bush, scientists kept finding antibodies from deadly diseases that had never spread: henipaviruses, sars, Ebola, “village outbreaks that are like flashes in a pan,” Relman said. On top of that, by bringing more humans into contact with wild animals, China’s vigorous wildlife trade had expanded the opportunities for such spillovers to occur. If that sounded a bit abstract, his second point in favor of zoonotic spillover was more concrete. By last summer, scientists had identified the closest known relatives of sars-CoV-2 in horseshoe bats. “The nearest known relatives of sars-CoV-2 are all found in bats, and they’re found in bats in China,” Relman said. “So you have to think at some point this virus or its immediate ancestors were found in bats—seems like a reasonable conclusion. The only question was: What was the path from bat to human?” In theory, at least, that isn’t such a complicated leap. “There are lots of cases where viruses go from bats to humans,” he said. But there were a number of reasons why the possibility of a lab leak remained viable. Chinese scientists have reportedly tested fifty thousand samples from three hundred species of wildlife, searching for the missing link, and hadn’t yet found one carrying sars-CoV-2. “The fact is, no one has found sars-CoV-2 anywhere other than in humans,” Relman said. “So that strikes me as a little odd.” At the same time, even in the United States, investigative reporters had found that “lab leaks are more common than you would hope.” The final data point that Relman mentioned was the one that is repeated most often: one of the very largest collections of bat coronaviruses was held in the Wuhan Institute of Virology, in the same city where the outbreak occurred. Meanwhile, the location where the closest known relatives of sars-CoV-2 were discovered is in Yunnan Province, which borders Myanmar, a thousand miles away... Baric said that sars-CoV-2 was different enough from known viruses that to retrofit it from an ancestral strain would have required a truly unprecedented feat of genetic reëngineering. “And of course you don’t know what you’re engineering, because sars-CoV-2 would not have existed,” Baric said. Another possibility was that a virus nearly identical to the eventual pathogen, which had been collected in the wild and stored in the virus repository, had somehow escaped containment, but he hadn’t seen any specific evidence to support that hypothesis, either.