An exhaustive, three-month trial to determine if nuclear weapons tests in Nevada caused dozens of cancer deaths ended today, with a government attorney arguing that the only connection between radioactive fallout and cancer deaths in the United States was in the minds of biased researchers.

U.S. Energy Department attorney Henry Gill, completing final arguments for a trial record that now exceeds 6,500 pages, suggested that a study showing unusual numbers of cancer deaths in southwestern Utah had been tailored to fit the claims of 1,192 relatives of cancer victims seeking as much as $250 million from the government.

The government's expert testimony, Gill said, showed that low-level radiation from atomic tests in the 1950s and 1960s could not be linked scientifically to the cancer cases.

U.S. District Court Judge Bruce S. Jenkins has indicated that he will take several weeks to reach a decision in the landmark case, which could open the way for many more suits if the government is found to be responsible for letting fallout drift into populated areas without adequate warning.

Lawyers for relatives of 24 alleged victims, only five of whom are still alive, said their expert testimony showed a clear connection between the tests and the cancers.

They recounted stories of children and parents dying painful deaths more than two decades after what they alleged to be a particularly uncontrolled test in 1953.

Tucson attorney Dale Haralson called the deaths "the greatest civilian tragedy that has ever been experienced by U.S. citizens since the Civil War."

Testimony in the case has indicated that officials at the Nevada test site often delayed bomb blasts when the wind threatened to carry fallout toward heavily populated Las Vegas and southern California, but proceeded with tests when the winds were blowing toward small Utah towns such as St. George, about 120 miles east of the test site.

Government attorneys argued today that the victims' relatives had waited too long after the cancers appeared to file claims against the government, under a law that they said put a two-year limit on such cases.

The government lawyers also argued that the financial information upon which damage claims were based was inaccurate and that, in any case, the government could not be sued for events arising from officials' using their discretionary policy-making powers protected by federal law.

Dr. Carl Johnson, a Denver physician and epidemiologist, testified during the trial that 288 cases of cancer had occurred in the affected area of southern Utah over two nine-year periods, when only 170 cases would have been expected in a similar-sized area.

The most prevalent cancers were those known to be caused by radiation. Leukemia occurred four times more often than expected and thyroid cancer six times more often than expected.

Johnson noted that above-ground testing in Nevada occurred from 1951 to 1963, and said that research on victims of the 1945 atomic bombings of Hiroshima and Nagasaki showed that cancers caused by radiation often take years to develop. "It seems probable that the largest increment of cancer in southwest Utah is yet to come," said a report prepared by Johnson.

Gill leveled his strongest charges at Johnson's study. Gill suggested the survey takers had been ill-trained, had not asked the right questions and may have been influenced by the knowledge that Utah residents were planning to sue the government over the cancer issue.

Gill called Johnson's testimony linking cancers in many of the 24 alleged victims to fallout as "incredible" because, he said, Johnson read only summaries of their cases and not their full medical records. Earlier in the trial, government attorneys also complained that Johnson's survey relied on the memories of cancer victims' relatives and was not backed up by medical records.

Ralph Hunsaker, a lawyer for the alleged victims, said the government knew by 1950 that Japanese atomic bombing victims had developed certain cancers and that it knew winds could carry debris from the blast sites to southern Utah. "They simply did not heed the dangers and they hoped to control the risks," he said.

Former U.S. interior secretary Stewart Udall, another attorney representing the Utah residents, contradicted the government's agreement that it is protected from suits by federal law.

Udall said the law applied only to high-level officials using their discretion to set broad policy. Lower-level officials at the test site had waived their right to claim immunity under the "good Samaritan" principle of law, Udall said, because they did make some effort to protect area residents from fallout, although the efforts were erratic and ineffective.

"When government officials decide to provide a service, they must do it with due care . . . and evenhandedly," Udall said. During the 1950 tests, he said, officials kept some cars out of dangerous areas and let others go through, warned some residents of nearby towns to wash themselves and discard their clothing but made no effort to give the same warning to others.

During one of the worst incidents in 1953, several of the attorneys pointed out today, a radiation cloud from a Nevada test passed over St. George, Utah, an hour before the town was warned.

"We're dealing with the most horrifying of injuries," said attorney Haralson. "We are dealing with slow, painful, agonizing and almost always terminal injuries. In many cases the treatment is worse than the injury itself."