That letter contaminated postal workers, equipment and thousands of pieces of mail.

Until recently, scientists thought passing the two-month mark meant leaving the danger zone. So far as scientists knew, anyone in danger of developing inhalational anthrax after breathing bacterial spores would show symptoms within 60 days.

But in the nine weeks since the first anthrax case was identified, scientists have had to reconsider much of what they thought they knew about the disease. Recognizing the gaps in its knowledge, the Centers for Disease Control and Prevention will shortly convene a group of academic experts to help develop an anthrax research agenda. The aim is to identify what can be said with confidence and what needs to be puzzled out.

The Atlanta-based CDC already has taken an unusual step. Thursday, it released suggestions on how those made nervous by the anthrax cases should handle their mail. In a departure from past practice, the agency emphasized that the suggestions were not "scientifically proven recommendations." The "don't list" included sniffing mail or bringing it near the face and ripping or blowing open envelopes.

"We're trying to balance protecting people in as absolute a way as we can from any unnecessary illness, while by the same token not creating undue alarm and fear where it is unwarranted," said Dr. Jeffrey Koplan, the CDC's director. "The risk of cross-contamination is very low, but there are some elements that we can't explain well."

One of the greatest remaining uncertainties is the degree to which contamination of the mail endangers third parties. The concern not only is about the politicians and media members who received the original letters or the postal workers in the vicinity while the letters were processed. How vulnerable are recipients of mail that passed through the same facilities as did the tainted letters?

The uncertainty persists because, after several weeks, the inhalational anthrax deaths of 61-year-old Kathy Nguyen of New York and 94-year-old Ottilie Lundgren of Connecticut remain unexplained. No anthrax has been found in their homes or other places they visited. But in Connecticut, a letter carrying a few bacterial spores was found in a nearby neighborhood, and in New York, a letter that went through contaminated machinery is believed to have been delivered nearby.

If a tiny number of anthrax spores could cause death, that would defy what is known about the bacterium. Determining whether it is possible, though, has CDC investigators following a trail similar to ones they have pursued before.

"It is not that unusual for us at CDC to have to understand something very technical in order to figure out what has happened with an outbreak," said Dr. Beth Bell, a medical epidemiologist who led the new Jersey investigation. "People who investigate [Legionnaire's disease] outbreaks have to learn about cooling towers and ventilation. When you investigate a waterborne disease outbreak, you have to understand how water systems and sewerage work. It is a fundamental part of our approach."

In this investigation, a half-dozen CDC personnel have had to become instant experts on postal machinery and the paths that letters take through equipment handling 500 pieces of mail a minute. Because a post office uses the same machinery for two different functions -- processing mail from its local area to send out elsewhere and sorting mail from other areas to deliver locally -- contamination from a single loaded letter could spread widely.

"At some point in this process, if aerosolization is occurring from a contaminated letter, your contaminating element is not a piece of equipment anymore; it's the air in which spores are now floating around," Koplan said.

Since Oct. 22, when cutaneous, or skin, anthrax was identified in a New Jersey woman who had not been to the post office, it has seemed likely that cross-contamination of mail could cause disease. But the CDC has not been able to link her to specific letters, and therefore has not been able to estimate what dose of spores she received.

"It was not possible to determine whether it was one letter, more than one letter or many letters that might have been the source," Bell said.

Some 22 days elapsed between the discovery of Nguyen's illness and Lundgren's. It has been 18 days since Lundgren's case was announced. Millions of pieces of mail passed through the contaminated post offices before they were closed. It is possible that anthrax spores still persist on some piece of mail and could cause another case. As the days pass, the CDC hopes it is unlikely.

"As an epidemiologist, I look at the numerator, and I look at the denominator, and I look at the quality of the information," Bell said. "As I look at this, I see an extremely small numerator of cases, and I see a very, very large denominator, both in terms of people who have received mail and in terms of sorted letters.