He called space shuttle engineers and asked his own questions. He marveled at the good sense -- and feeble clout -- of this agency he had never heard of, the GAO. He made his own, surreptitious investigation of possible White House pressure to launch the ill-fated Challenger.

Professor Feynman went to Washington, and as he does in most places, he had a very interesting time. But the boyish 68-year-old Nobel laureate concluded today that it was a place sadly awash in public relations -- a motivating force he said was beyond his understanding yet at the root of the Jan. 28 shuttle explosion.

California Institute of Technology physicist Richard P. Feynman, back in his beloved California today after his labors as a member of the blue-ribbon presidential commission on the shuttle disaster, invited the press to campus to hear his story of adventures in the East. It was a mix of statistical analysis, scientific caution and mischievous speculation, climaxed by his conclusion that somewhere, somehow, some top National Aeronautics and Space Administration official approaching a congressional hearing may have said to a doubt-ridden engineer, "Just be quiet. We're trying to get money."

Feynman bristled at a NASA contention that there was a mere 1-in-100,000 chance of a total shuttle flight disaster. In 13 pages of "personal observations" he released today and scheduled to be printed in the commission report appendix, Feynman noted that an analysis by NASA engineers of previous rocket flights showed a 1-in-100 chance far more realistic.

He said NASA official Judson Lovingood, now associate engineering director at the Marshall Space Flight Center, told him that the 1-in-100,000 estimate was based on "engineering judgment." "As far as I could tell," Feynman said, "the engineering judgment, so called, consisted of making up numbers."

Feynman drew attention early in the commission hearings by demonstrating with a glass of ice water how the rubbery material used to seal joints in the shuttle's solid fuel rockets might have been dangerously stiffened by cold weather the night before the launch. His report said NASA clung to optimistic estimates of the shuttle's capabilities despite evidence of erosion of the material in previous successful flights. "When playing Russian roulette," he said, "the fact that the first shot got off safely is little comfort for the next."

He said he initially refused to sign the commission report because he opposed a draft recommendation that the group "strongly recommend" continued support for NASA and the space program. He said he did not think the commission had discussed this recommendation, but agreed to sign after the recommendation was downgraded to a "closing thought" and the language changed to "urging" support for the space agency.

Pressed by reporters, Feynman said he thought a reasonably safe shuttle vehicle could be built, but only if NASA has sufficient time and money. The best disaster ratio the agency could hope for, he said, was one failure in every 1,000 flights. He grumbled at being pinned down on such a figure. "It's a dumb answer," he said.

Winner of the 1965 Nobel Prize for physics and author of a recent best-selling, off-beat memoir, Feynman spoke of Washington as alien turf often inexplicable to "a man from an ivory tower." Asked how to get good independent evaluations of NASA's work, Feynman said, "I may be naive here, and I don't know much about government . . . but I found that there is an organization that is doing a tremendous amount of work in evaluating things and [his voice dropped to a stage whisper] as far as I can tell nobody pays any attention to it . . . . It's called the GAO or something like that, the General whatever-it-is." He was referring to the General Accounting Office, the investigative arm of Congress.

Feynman, calling himself a "very suspicious character," said he conducted a one-man probe into possible White House pressure to force a Jan. 28 launch. "I feel sorry about this, because it was one of the few times that I did something surreptitiously," he said.

His hypothesis was that teacher Christa McAuliffe's planned broadcast from space, a pet White House project, might have required days of technical preparation. A chat with technicians might expose premature orders from administration officials determined to get it done on schedule. "So I went and investigated the telemetry systems both at Kennedy and Johnson [space centers]," he said. Friendly technicians, however, quickly informed him that no great preparations or White House orders were necessary because it would take them only three minutes "to set the switches" for a broadcast.

Still, Feynman said, NASA's habit of discouraging talk of problems in order to prevent funding cutbacks may have had a large impact on the disaster. "Good bureaucrats know what to do without being told," he said.

He pleaded for more realistic planning at NASA. "Dreams do not have to be built on fantasy," he said. "Dreams are built best on reality plus a little bit."

When appointed to the commission, Feynman said he would be happy to ride the shuttle some day. Today he amended that. A ride in space "could be interesting," he said, but "I would have a large amount of thinking to do before I made up my mind."