Nine billion pounds of meat, tons of fowl, fish, cheese and even pet food -- on one day in August of 1978 all of it became suspected poison. Ten percent of the entire food supply of America.

Sodium nitrite -- the salt that is used to preserve, color and flavor our foods -- apparently causes cancer, two regulatory agencies announced jointly on that day.

Nitrite soon became both famous and feared. The hot dog, symbol of the casual American diet, seemed suddenly tainted, colored with disease. And not any diease, but the disease that is unpredictable, lingering, disfiguring and extremely painful. Cancer, according to public opinion polls, is something Americans fear more than war.

Then, one day last fall, with what seemed to be equal suddenness, the regulatory agencies jointly said, "Never mind." It was almost two years to the day after their warnings that the regulators said they couldn't prove that nitrite causes cancer after all. "I had a hot dog for lunch," crowed the commissioner of the Food and Drug Administration.

What happened with nitrites is a classic story of the combat between public and private interests, between difficult science and irrascible politics. It is a struggle the regulator -- no matter what his political affiliation -- must face daily: Decisions must be based on science, which is always slow and often ambiguous, despite the pressures of Washington politics, which are quick, clear and powerful.

That is a volatile mixture. If the regulator goes slowly, as the scientific results are checked and rechecked to assure accuracy, he can be accused of covering up what he already thinks he knows, of being a tool of the business interests he is supposed to regulate, of fiddling while the public burns. If he moves too quickly, he can be accused of recklessly toying with people's jobs, of needlessly damaging industry and the economy, of becoming that government shibboleth -- the Washington regulator run amok.

One incident in the nitrite episode expressed the dilemma well. Donald Kennedy, then head of the Food and Drug Administration, wrote in defense of his move to ban nitrites gradually: "Scientific peer review tries to establish the validity of a test . . . It is not a complete guarantee of being right . . . [but] regulators can and must be willing to proceed with less than this guarantee. When serious health risks are involved, scientific evidence suggestive of these risks should be enough to justify regulatory action."

(Someone in the meat industry got hold of this internal working paper and marked in the margin: "BIG MISTAKE!")

Later in the same document, Kennedy concludes, "There are special reasons why we should [run risks of this sort] when the public health is at stake. You have a politically expendable regulator [Kennedy himself] who is prepared to take the blame and will do so if the outcome proves more disastrous than this argument predicts."

At the beginning of the nitrite case, when Kennedy realized how difficult regulation of this simple salt would be, he told his colleagues that this case would be the real, best test of them as regulators.

The test may have been too hard. The result appeared to outsiders to be a series of chaotic charges, and then retreat in disarray.

The small laboratory building in Boston, its bricks blackened with age and soot, was the place the whole drama began. On the top floor is the office of Dr. Paul M. Newberne, a mild-looking man with gray hair and an easy manner.

Newberne is a specialist in animal diet and disease at Massachusetts Institute of Technology. From 1976 to 1978, he raised about 2,100 rats in his laboratory, and fed them nitrite in a dozen different ways. He was testing, for the FDA, the proposition that nitrites cause cancer in animals.

By the winter of 1978, the aniamls were dead -- of cancer, of natural causes, or at the hands of researchers anxious to finish up the experiment and get the cells of the last aged stragglers onto miscroscope slides.

Lab assistants took the tiny organs from the bodies, picked them in preservative, inked them with blue stain, impressed them in wax, then sliced them 7/1000 of an inch thick. With the features darkened and the tissue now thinner than paper, the animals' cells had become translucent.

Newberne bent over his gray, double-eyepiece microscope and painstakingly began to count. Among the circular globs and granules of the cells were ones that had become elephantine. There were cells stunted and gnarled with disease. He named cell types and made diagnoses out loud as a secretary wrote. It was a long list that was taken down. There were 50,000 slides to see, to recognize, to judge and to diagnose.

What Paul Newberne found surprised him. He did not simply send a report to the FDA. Nor did he even call in a report. He decided he had better carry his news to Washington in person.

Mankind has been curing meats with salts for 10,000 years without benefit of regulatory agencies to guard the public health. There are cave paintings that show meat being cured with sea salt, a good source of nitrite.

But it was not until this century that chemists discovered that the active disinfectant that prevents deadly botulism poisoning is sodium nitrite. We now know that nitrite is virtually everywhere in nature. It is in vegetables in sizable amounts. It is manufactured in human saliva constantly. In fact, our intake of nitrite from cured products is only a small percentage of our total intake of the chemical.

The hazards of nitrite were discovered beginning in 1954 when scientists began to learn that nitrites, under certain conditions, can combine with a variety of chemicals called amines. Together they make nitrosamines, among the most potent cancer-causing agents known to man. Nitrosamines have been found to be carcinogens in 22 species of animals -- a deadly superlative among chemicals.

In 1971, the Food and Drug Administration contracted with Newberne and a colleague to study nitrosamines. Newberne got the expected deadly result from his animals. But he got something else as well, something far more disturbing. In one control group, rats had been given nitrite alone, with no amines. Newberne expected to see no nitrosamines formed and no cancers following.

But instead, Newberne reported, this group showed a startling number of lymphomas -- malignant tumors, cancers of the lymph system.

It was still two years before Newberne would begin his big study of nitrite, using 50,000 slides. This result was only in a side group, an accidental finding, and could not be relied on for scientific accuracy. But still, it was startling.

The findings meant that nitrite itself might be a carcinogen. If that were found to be true, then the whole focus of worry would have to be shifted back one level, from the nitrosamines to the nitrite itself. Worse, food manufacturers were not adding nitrosamines to food, but were adding nitrite to billions of pounds of meat, poultry, cheese and fish.

So it was a few dozen rats, and Paul Newberne's diagnosis of their lymphomas, that began the early history of the great nitrite scare. As a result, the FDA pushed Newberne to do a monumentally large study to learn whether nitrites themselves could cause cancer.

Animal studies of cancer had not been done in any numbers until a few years earlier, in the late 1960s. "The diagnoses in those dark ages went unquestioned," said Dr. Robert Squire of The Johns Hopkins University. "Everybody believed everyone else. The Good Laboratory Practices standard [a uniform standard to assure consistent quality in lab work] hadn't been established yet, and no one suspected the kind of irregularities that could occur."

It is a typical irony in this tale of nitrite that the first accidental result now appears to have been misread and mistaken, according to the pathologist who reviewed the old slides for the government.

"If this early study had been looked at the way we look at these things today, there would have been no need for everything that has happened since . . . I found that half the things called lymphomas were not lymphomas. There was nothing there" to set off the alarm on nitrite, said Squire, who reviewed the work for the government.

Gary Flamm, then assistant commissioner of FDA, said: "Of course, Squire may be right . . . We didn't review that study, but if it came in now I certainly think we would review it in some detail. God, yes."

But the FDA didn't review it. Instead, a much larger study was commissioned to look at nitrite alone, to try to prove or disprove that nitrite itself can cause cancer in rats. That study began in 1975.

There were more than a hundred rats in each of 18 groups, each group fed a slightly different way. Some were fed nitrite in their rat chow, some were fed nitrite in their water. Some got large doses, some smaller doses and some none at all. For Newberne, the result was the 50,000 slides to examine every day for months.

Pathology -- the examination of tissue to find abnormalities -- is notoriously tricky. The pathologist must be familiar with the look of each different sort of tissue as well as the scores of different tumor types, both malignant and benign.

Paul Newberne, a tall man with a soft southern voice, peers into the gray microscope near his desk and says, "If you've got 10 years of experience looking at that under all sorts of conditions, you know whether it is normal or not. You get a sense of how distended the parts are, whether they are irregular or too long for their type, even though only a few of them are actually misshapen."

Newberne started his career 25 years ago as a veterinarian, working for three years in rural Georgia before returning to Auburn University for his Ph.D. in Pathology and the beginning of his animal research center. Along the way, Newberne had earned two distinguished fellowships to study animal pathology, and from school he went directly to head a program in nutritional pathology at Auburn.

In the past two decades, he has continued to devote himself to the study of diet and its effect in animal disease.

"This is why we don't all agree on tumors," said Newberne, nodding toward the microscope. What was there looked like the spotted surface of granite. "The tumors show up at such a wide range in their stages of development, and they look different at each stage. The question is whether it has advanced to the stage where it is irreversible, malignant and must be removed before it kills the animal."

By spring of 1978, Newberne had eyed his 50,000 slides of tissue. Seeing the slides one at a time, he had no sense of their pattern; he did not think he had found any special effect. It was only when he tabulated the number of tumors that he began to think he was finding groups of nitrite-fed mice with two to four times more lymphomas.

Newberne was surprised and a little excited. He flew to Washington to meet with Donald Kennedy, then commissioner of the FDA.

"At that meeting we pushed Newberne very hard on a variety of questions," said Kennedy. "I thought he gave a strong account of himself, and the conclusion at the end of the grilling of a couple of hours was that a report would soon be in the public record as a major indictment of nitrite."

Newberne recalls showing his data, scratched out on a big pad of paper, to a roomful of people. "I said the result was only suggestive and we needed more studies to find out what's going on. That's all I said. I didn't tell them to go ban nitrite. Never."

It was the beginning of May 1978, and the news spread like a shock wave among the small number of people allowed to hear it. Donald Kennedy was upset, according to those who talked with him later. The FDA had just come through a loud, bitter public dispute over the banning of saccharin, the largest public battle the FDA had ever had to face, bigger than the cyclamate case, bigger than the cranberry brouhaha of years past.

Then, up popped an issue much larger and much more complex. From the beginning Kennedy told his colleagues that this case would be their best test as regulators. It was legally complicated because nitrite was governed by several different, competing laws. It was politically scary because there were possibly billions of dollars worth of consequences that would come from any decision. It was a public health nightmare because it was not a simple matter of removing a hazard from the market -- the regulators had to balance two hazards. How many fatal cases of botulism or other disease would occur if nitrite were banned as a protector of foods? How many cases of cancer would be caused if it were not banned?

Kennedy called Carol Tucker Foreman, the assistant secretary of agriculture in charge of food quality and consumer services. Her activist approach to food and health issues had already earned her the title of "dragon lady" among pork producers of the Midwest. They believed that "whenever she opened her mouth, pork prices dropped," one meat industry spokesman said.

"Apparently it's true," Foreman said, "that whenever the Commodity News Service carried something on the issue of nitrosamines and our department, the pork belly prices fell the limit. There was quite a lot of fear engendered among the farmers. We just don't know how, in a chemical-filled world, not to frighten people from time to time as we deal with these issues."

Foreman had been wrestling for a few years with the problem of nitrosamines found in bacon. At just about the time Don Kennedy called, she thought she had finally wrestled it to the mat, arriving at a technical compromise with the meat industry.

She answered the phone in her large second-floor office. "I have some very bad news for you," Kennedy told her. He said he had a study showing nitrite could cause cancer directly, never mind the nitrosamines.

"Oh s---, Foreman said, and sank through the same rapid, dismal thoughts that other regulators say they also had when they heard the news. "All this struggle over the nitrosamines, and now this guy comes in and says that nitrites themselves are the bad actors. This was obviously something much bigger, much worse."

Commissioner Kennedy gathered a group of half a dozen top agency officials to decide what must be done. Contrary to the procedure on smaller matters, he did not turn the Newberne study over to his agency's senior scientists. Kennedy knew that the Newberne report contained political dynamite. He also knew that the FDA was a sieve, which constantly leaked to the press. The fewer people who knew of this report, he figured, the more likely his group would be able to work in peace. This would later turn out to be a key mistake.

Aware that scientific review would take place at some point, the Kennedy group simply went to work on the assumption that Newberne was right. Newberne, after all, had strong reputation. He was believed to be an especially cautious man in his work and what he said about it.

A single thought hung over the first meetings like a pall: Nowhere had man actually faced the problem of preserving meat without salt. The problem was new and on a scale quite beyond the saccharin flap that had so occupied the previous two years. Epidemics of botulism poisoning could hit the nation if nitrite were suddenly banned. And the applicable law -- called the Delaney amendment -- seemed to demand an outright ban with no phaseout period.

It was May, and the group talked of picnics: families taking luncheon meat or sausage in plastic wrap and sealed containers. The fatal botulin bacteria grows rapidly in the absence of air. The inside of a sausage or a hot dog is a good place for it to breed. They envisioned children carrying their lunch boxed to school, setting them in the sun for a few hours before lunchtime. That evening, some of them might not come home.

"Since we had no experience with the situation, we had no idea what the real risks were, if some other infections might not result as well," said Flamm. "After all, people will do anything with a sandwich. They will subject a sandwich to the most incredible conditions -- one man will take it to a foundry and set it down in better than a hundred-degree heat, another who drives a refrigerator truck will put it back there and let it freeze before lunchtime," said Flamm.

The problems facing Kennedy's caucus were more than the fear of disease. Since man has used nitrite salts for so long in so many ways, its trail through the law is a tangle.

When Kennedy's caucus sorted it out in 1978, it was clear that the law was very strict: It demanded that when a substance is found harmful or carcinogenic, it must be banned.

The law simply allowed no phase-out, moratorium or other ways of backing into regulation. In the case of nitrite, Kennedy and Foreman knew a public health disaster was possible if botulism and food poisoning were not carefully contained after a ban of nitrite.

A nitrite ban could lead to economic chaos for the meat industry, which in nitrite-treated red meat alone is a $12.5 billion operation.

By law, the regulators simply could not consider these things. They did anyway.

Kennedy's strategy was to develop a 50-page paper detailing the options ranging from no action at all to an immediate ban, and then choosing the most reasonable course -- a phase-out of nitrite based on the speed with which the nation could find substitute ways of preserving meat.

The group found that refrigeration would not work -- too few refrigerators, too many that might break down and let the meat temperature rise above the 40-degree safety mark. Luncheon meat, hot dogs and the like would have to be frozen. To do that, the nation's freezer capacity would have to be doubled.

Over the years, about 700 chemicals have been tested to see if any might make acceptable substitutes. Nearly all have been failures.

To get around the rigid law, Kennedy and his group decided to adopt the tactics of their opposition in the saccharin battle. In that one, they were forced to give up their ban when Congress passed a law declaring a moratorium on the issue. Kennedy thought they might bring their own bill to Congress -- this time to prevent themselves from banning a substance.

Up to this time, Kennedy had not worried about the scientific data. He had sent the Newberne report to reputable scientists around the country for them to review, checking statistics and logic. Of course, they could not check the pathology. The reports came back positive -- the paper itself seemed all right.

At that moment, in late June, Kennedy had no idea that the Newberne study would eventually go well beyond the usual peer review that scientific papers get. "There are two kinds of scientific review," Kennedy explained. "One is peer review, in which a paper is passed out to referees to look it over and see if the conclusions can be justified by the data. A pathology review is something entirely different. It is like a new experiment, only it takes as raw material the data of an older experiment. It is costly and time-consuming and you don't conduct one until it becomes clear that there is a problem with the study.

"Early on, it was not apparent to me that a pathology review was necessary. Actually, earliest on, I wouldn't have known a pathology report if it ran over me in the street," Kennedy said.

"What one does as commissioner when you've got issues in half a dozen scientific areas buzzing around you -- you just ask the best questions you can, and decide on the basis of the answers what needs to be done."

So he and his small group of officials worked out a compromise among bitterly competing interests. Only later, at the insistence of other scientists at the FDA, was the deeper scientific review started. A little too late to avoid embarrassment, as it turned out.

Outside the closed circle of Kennedy's caucus, those other FDA scientists had been worried.Politics was being discussed and decided, but no science. The senior scientists waited through the summer, but were not consulted and were not given copies of Newberne's study to review.

The action by Kennedy to keep Newberne's study secret even from the senior scientists at the FDA is the decision that has since been most vigorously attacked by critics. It has also been criticized in a General Accounting Office investigation and in a Congressional Research Service report.

After the FDA had Newberne's report for two months, chief scientists at the FDA began to realize that they were being cut out of the decision-making. Whether their judgment was necessary or even useful, they felt hurt to be left out, and upset that their skeptical point of view was not getting a proper hearing. No scientific review was taking place. But political plans were already far along.

"On many an occasion," said Dr. Albert Kolbye, associate science director of the Fooods Division of the FDA, "I expressed to Dr. [Howard] Roberts, my immediate superior, my growing dismay and concern that events were proceeding too quickly without an adequate foundation of scientific review. I have never experienced before in my government career a situation where the word was [that] most of us were to be excluded and not to get involved."

Flamm, too, worried. "We had been alerted to something that was wrong with the study. Newberne had called in and changed the numbers of rats supposed to have tumors. He changed them not once, but twice, then three times that I heard about. And they all moved the number of tumors downward. This is extraordinary, and it make us very nervous.

"This was an important case," Flamm said. "I was concerned, extremely concerned, that I should do the right thing . . . During that time I began to walk in the woods of West Virginia with my wife to talk about it. I knew that if I screwed up on this, I would leave government work altogether.

"When the Newberne numbers started to change, we told Kennedy that it was series. Our case might be going straight down the tube."

Then, the scientists heard that Kennedy was likely to hold a press conference soon to announce plans. They heard the news first by rumor and then by reading it in Food and Chemical News, a trade publication. Kolbye barked loudly about this. Before the supposed press conference, Gary Flamm was asked to help anticipate critical questions about the Newberne study.

"I was role-playing. I was supposed to be Giovanni Salami, the head of the Meat Industry of America. I worked over two FDA scientists and criticized the Newberne report. I think I hit it impressively hard," Flamm said.

Kennedy says he was not told of the three successive changes in Newberne's data. If he had been, he says, it would have shocked him into some quick action. But by the time Kennedy turned from the political and legal questions to the scientific ones, it was August.

Suddenly, the bubble of privacy enjoyed at FDA was burst.

Marian Burros, food editor of The Washington Post, called the FDA, saying she knew of the Newberne study and was going to run a story on it. The FDA concluded that a public statement had to be made.

"Nothing was secure at FDA," Kennedy lamented. "I actually had members of Congress getting internal memos before I got them!"

On a Friday, Aug. 11, a joint statement was issued by the FDA and the USDA.

"A study recently completed for FDA by the Massachusetts Institute of Technology strongly suggests that nitrite produces cancer of the lymphatic system in test animals," the statement said.

Within a few days the 50-page document that the Kennedy caucus had been working on was in the hands of the press. The report said: "FDA has received a report from Dr. Paul Newberne of MIT. . . The agency has evaluated the study and has concluded that it shows that nitrate induces cancer when ingested in laboratory rats. . . FDA expects that its conclusion will withstand further analysis . . . Accordingly, FDA and USDA are planning for a phaseout of nitrate over a period of several years." u

There was no more time for deliberation. Everyone was in the water now; the free-for-all began.

The loudest squeal came from the pork producers. "Consumerist zeal running roughshod over science and common sense," was one observation. "The meat industry," said a Senate Agriculture Committee report, "especially the pork producers, are convinced that removal of nitrate will result in their economic ruin."

One of the unspoken concerns at the time, but one which may have been a major factor in the thinking of meat industry executives, was mentioned by Paul Newberne: "You have to remember that nitrites let them off the hook on many points.Not just botulism. They can be a lot less sanitary in their plants, and so they can spend a lot less money cleaning up and using special equipment. Nitrite, you know is an excellent disinfectant. I suspect this thought played an important role."

"A nitrate ban would cause prepared meats to disappear," said a Tennessee newspaper in quoting a meat industry spokesman, "disrupting the food supply. The hog industry would be decimated, since 70 percent of U.S. pork ends up in processed meats. What is more. . . nitrate protects against the poison that causes botulism. Banning nitrite could save, say, 2,000 lives from cancer each year and lose 10,000 to food poisoning."

Consumer activists were not happy either. "Carol Foreman and Donald Kennedy should go back to the law and read it because they are violating it," said Ralph Nader. The Ednvironmental Defense Fund attacked the agencies for "laying low when they should be racing ahead."

Although the consumerists wanted an immediate ban, the industry wanted no ban and no hints of severe action either.

But even as the FDA and the USDA were weaving a defense for their actions on TV and in interviews, FDA scientists were unraveling it from the other end.

Within six weeks of the announcement that would have altered the eating habits of virtually the entire nation, FDA scientists already knew that the Newberne study had problems. In late September, two independent pathologists looked at the slides that were supposed to contain the rat tumors found by Newberne. The two worked separately, but their conclusions were virtually identical.

They disagreed with Newberne on 29 of 35 animals checked. This result was enough to knock the Newberne study below the level of statistical significance, even though only a seventh of the tumorous animals had been checked.

By today's standards, Newberne's diagnoses of that time are considered loose. He lumped together different sorts of tumors under the category of lymphoma.

The FDA soon decided to have all 50,000 slides done by Newberne reviewed by other expert pathologists -- a group called the Universities Associated for Research and Education in Pathology (UAREP). Forty-six pathologists looked at all the slides. Their conclusions were just as negative as the earlier reports:

Of the 2,100 or so animals studied, the reviewing pathologists disagreed with Newberne in 897 cases, an enormous 43 percent disagreement. The reviewers normally expect disagreement of only a few percent.

Of the 220 or so tumors that Newberne called cancers of the lymph system -- and this was the heart of the study, the reason for banning nitrite -- the reviewers could find only 30 such tumors. This is only 14 percent of what Newberne claimed.

"We are at a loss to understand this enormous difference," said Kenneth Endicott, chief of UAREP. Some pathologists have believed that it could not have been Newberne himself who did the diagnoses, that he must have allowed an inexperienced student to do it.

Newberne says he did it himself. He also disagrees with the diagnoses of the many pathologists who reviewed the study.

Says Newberne: "I may be wrong in my diagnoses. All right. But whatever I call it, something is there, and it is not good. I don't give a damn whether you call it lymphoma or call it what the UAREP people call it, histiocytic sarcoma. That doesn't make a damn bit of difference to me. They are malignant tumors. That's what counts."

UAREP's Endicott says, "It does matter. Our pathologists believed the tumors were not derived from the lymph system, but originated in the connective tissue. It is like having an assortment of fruit; when you compare them, do you separate out the oranges and the grapefruits?"

Since the study as a whole identified about two dozen different tumors in various organs, the categories to which the tumors are assigned do matter. If many sorts of tumors are grouped together, the category in which they have been placed might become large enough to indicate a significant result for the study as a whole.

Why Newberne would end up looking so mistaken when his previous work and his reputation have been so strong is "a puzzle that none of us have been able to solve," said one pathologist. "It is really intriguing, especially when you go back to the first small study, the one that started all this. The diagnoses there had exactly the same problems. This is not something a pathologist of Newberne's experience would do. So why did he? No one knows."

"Look," said Kennedy, "science is conducted on the assumption that when you have a seasoned, well-respected investigator from a good institution, and he reaches a scientific conclusion, it is far more likely to be right than wrong.

"In this case you can certainly conclude that Newberne's judgment on what you call a tumor is different from other experts' in the field. It is clear Newberne was doing something very different, but not necessarily wrong."

Nitrite was found not to be a direct cause of cancer, the FDA concluded, though some doubt remains. Donald Kennedy thinks that something suspicious may exist in lower doses ranges that showed abnormalities, but no tumors. And untouched by the episode altogether is the fact that nitrites do sometimes become potent carcinogens as nitrosamines, though the process is not well understood.

But in the Newberne case, two years and seven days after the bad news on nitrite was first announced, it was repealed.

Smug remarks could be heard from all quarters as the biggest regulatory case ever belly landed, then fell to pieces, leaving behind a heap of smoking wires, wings and struts.

The pork industry says the cost of the whole episode could hardly be calculated, but if someone tries, it says, one should begin with the billion-dollar loss of sales in pork bellies. Though others disbelieve it, the pork producers say the losses occurred when millions of people conscientiously cut bacon and hot dogs from their diets in response to the nitrite "scare."

But the biggest cost of the whole episode, as one congressional staff said, may have been to the public's credulity:

"To cry wolf as big and as loud as it was yelled in this case -- that's enough to make us all start giving these agencies some funny looks next time they start talking about what causes cancer. I don't know if that's good or that's bad. . ."