A few details of Fouchier’s experiment were released last month, allaying some fears about its hazards. Kawaoka revealed an even fuller version of his work here Tuesday, further defusing worries. Numerous listeners, however, said it is only a matter of time before the question of whether to publish the results of “dual-use research” — research that could be used for good or bad purposes — comes up again.

Normally, bird flu is hard for people to catch. It requires close contact with sick birds and almost never passes from person to person. The ease of transmission is mostly determined by the structure of one protein, hemagglutinin. Kawaoka wanted to find out what mutations in that protein’s gene might make the virus more contagious in people.

He put a bird flu hemagglutinin gene into the 2009 pandemic “swine flu” virus and by various methods induced four mutations in it. The final bug was easily passed between ferrets, unlike viruses containing a “wild” bird flu hemagglutinin gene. But the engineered virus didn’t kill the animals and didn’t even make them as sick as the swine flu virus. The infections were also easily stopped with the drug Tamiflu.

In short, it wasn’t a Doomsday virus, as rumors about Kawaoka’s and Fouchier’s work had suggested. It was, however, a useful one, Kawaoka argued to the Royal Society audience of about 250.

Three of the four mutations were near a part of the hemagglutinin protein that attaches to cells of the nose, throat and lungs. H5N1 flu samples from Egypt are more likely to have changes in that part of the protein than samples found elsewhere.In the past three years, 71 percent of Egyptian samples from chickens, and all the samples from human cases, have mutations in that site.

This suggests that Egypt’s bird flu has a particular propensity for human infection and may warrant especially close surveillance. “This information is important for risk assessment,” Kawaoka said.

Fouchier’s experiment started with a whole H5N1 bird flu virus that came from Indonesia, which has had more human cases than any other country. He then genetically doctored it in a way that he is not allowed to describe because the Dutch government has put an “export control” on the information. At a meeting in Washington in March, however, Fouchier revealed that the virus killed ferrets only if it was sprayed directly into their windpipes. When it passed cage to cage in the air as the animals coughed and sneezed, none died.

“We have heard in the press that if this escapes, it will kill half the world’s population. I find that very doubtful,” he said here Tuesday.

However, many people who read the manuscript Fouchier provided to the biosecurity panel said it was impossible to tell how lethal his virus was.

“It was not clear. It was not clear,” said Robert G. Webster, a flu researcher from St. Jude Children’s Research Hospital in Memphis whom the committee called on for advice. He added, “There was a great misunderstanding on my own part, and on the part of the [committee], on the question of a transmissible lethal virus in ferrets.”

Fouchier has now cleared that up.

Staff writer David Brown is one of the speakers at the Royal Society meeting.