For 11 years, they sat in racks in a freezer in Atlanta - 28 vials, each about 2 1/2 inches high and a half inch in diameter, each bearing about a half inch of dark yellowish ice.

Numbers 8199, 8202, 8204, 8207, 8213, 8216 and the 22 others. They were all the evidence that remained of a mystery disease that swept St. Elizabeths Hospital in July and August, 1965, killing 16 patients and infecting 78 others before running its course.

When scientists at the Center for Disease Control in Atlanta isolated a previously unknown bacteria as the cause of "Legion fever," the disease that claimed 29 lives in Philadelphia last summer, they called for the 28 vials.

Left undisturbed at room temperature from one to four hours, each of the little chips of ice became what it was in the summer of 1965 - the serum, or liquid component of the blood of a survivor of "St. Elizabeths fever." The vials were paired into 14 sets, each containing a sample of the serum taken from the patient when he or she was the sickest, and a sample taken during the patient's recovery period.

Dr. Joseph McDade, the CDC lab scientist who finally isolated the Philadelphia bacteria, introduced some of that bacteria, introduced some of that bacteria into each of the St. Elizabeths' sera samples and then tested the samples to see whether they contained antibodies to the bacteria.

When blood is invaded by a foreign protein, such as a bacterium, it produces antibodies - agents to fight the bacterium. If that same bacterium is re-introduced into a sample of the same blood, and there is an observable reaction of antibodies, that means simply that the same bacterium had been present earlier.

Thirteen of the 14 pairs - which Dr. Albert Balows, director of the bacteriology section of CDC's laboratory division, calls "an amazingly high" result - contained the antibodies, indicating that the Philadelphia bacterium is the same micro organism that caused the St. Elizabeths' outbreak.

Some CDC investigators had suspected since it first appeared last August that "Philadelphia fever" and "St. Elizabeths fever" were caused by the same agent. Many of the victims developed penumonia accompanied by a high fever. In each case the patient experienced difficulty in walking and developed a dry cough. And in each case there was no explainable cause for the disease.

"John Bennett, director of the (CDC) bacterial disease section, said he felt we would not only solve this outbreak but his outbreak," said Dr. David W. Fraser, CDC's chief epidemiologist - disease hunter - on the scene in Philadelphia. The St. Elizabeths fever was Bennett's "first outbreak" while at CDC, Fraser said.

Although it was never determined how the St. Elizabeths fever spread, "it was thought to be airborne," Fraser said, "and there's no reason to change the suspicion. It fits the data. We've thought from the beginning that the Phailadelphia problem might be airborn."

It was an obvious step, then, for the researchers to turn to the CDC serum bank in their attempts to find evidence of previous outbreaks of the Philadelphia disease.

The so-called bank is a "huge room with several upright deep-freezes," said Balows. "They do not frost up and are designed to hold the racks of serum."

Each rack in the freezers is filled with paired vials of serum, "sealed with a rubber stopper and an aluminum cap. The vials, said Balows, "sit there just like soldiers in a row, one behind another."

Balows said the bank also contains sera drawn from persons who had been immunized against various diseases, and persons involved in outbreaks whose cause the CDC determined. Both those types of sera are kept for use as controls in future studies, for educational purposes and for comparison with sera taken from victims of diseases whose origins have not been traced.

The solution of the first major part of the Philadelphia puzzle - and with it that at St. Elizabeths - leaves only one mass mystery frozen in the CDC serum bank - the outbreak of 1968 in Pontiac, Mich.

"Pontiac fever," as it has become known, also was a pneumonia, and a pneumonia whose cause has remained a mystery. Although Philadelphia and St. Elizabeths fevers also are pneumonias, CDC scientists say they do not believe Pontiac fever is the same disease.

"There are some single individuals" whose diseases were never diagnosed and whose sera rest in the freezers, said Balows. "But the vast majority of (serum) represents known, rather than unknown, quantities," he said. "I would say they go back to the late 1940s or early 1950s."

Although scientists now know what killed the 29 persons in Philadelphia and the 16 at St. Elizabeths, they still have no idea where the bacterium came from, how it developed, how it spread to where it might turn up again.

Asked if it was possible that the bacterium is responsible for countless causes of pneumonia every year, Fraser replied, "I don't think we have a feeling for how frequently it pops up. A lot of things could turn out, but we just don't know. I'll be interest to find out."

Fraser said that even if studies had been done in the past on what might be called "normal" pneumonia victims, the bacterium might not have been noticed. "I think it would be unusual for this to turn up in a typical university hospital laboratory," said Fraser.

Asked when he will have more answers, Fraser said: "Maybe 10 years. Maybe never. Maybe a couple of months. I just don't know."