Seven of the 12 rubbery O-ring seals used to join segments of the Challenger space shuttle's booster rockets were not properly inspected before the fatal Jan. 28 explosion, including the seal at the joint where hot gases leaked, a space agency official told the commission investigating the disaster today.

National Aeronautics and Space Administration officials also offered evidence today that the same joint may have been damaged before the flight, possibly during significant difficulties encountered in connecting the booster segments properly.

Finally, a photograph displayed today to the presidential panel showed a small but significant irregularity, a gap between the O-ring and the metal booster segment, at roughly the same location.

The convergence of these problems at one place on the booster took on even greater importance at the hearings today when NASA officials released computer-enhanced photographs pinpointing the fiery plume that preceded the explosion more precisely than before. It was at the same joint where the other irregularities have been discovered.

Officials did not say that any one or all of these factors caused the explosion, but said they were among a half-dozen new theories or "failure scenarios" that are now being pursued. These theories, which range from mistakes in assembly of the rockets to the presence of rain water and ice on the launch pad, suggest a broader scope to the inquiry than has previously been disclosed.

While the O-rings have long been suspected, NASA officials presented new evidence that they said indicated that design or manufacturing defects could have accounted for the seals' failure rather than the effects of cold weather, as previously thought.

"In reviewing records of inspections, we found that seven out of the 12 O-rings were not subjected to the same level of inspections as previous O-rings had been," said Jerrol Littles, associate director of engineering at the Marshall Space Flight Center in Huntsville, Ala.

Littles said that because of an "engineering-change order" on the O-rings, the seven seals used on Challenger's rockets were inspected by officials at Morton Thiokol Inc., the rocket booster maker, but not by backup Air Force inspectors, as was required, Littles said.

An Air Force spokeswoman said last night that in March 1985, Morton Thiokol changed the engineering drawings for the O-rings and "without government approval, their quality engineering division deleted the inspection criteria" for seals the firm purchased from O-ring manufacturer, Hydrapak Inc. of Salt Lake City.

The spokeswoman said this resulted "in an inadvertent deletion of three mandatory inspections for the O-rings," but added that inspections were done by either Morton Thiokol or Hydrapak.

The testimony on the inspections came after Littles showed an enhanced photograph of one of the O-ring seals that he said showed a "gap" that would allow hot gases from the rocket to "blow by" the seals.

"I believe it is realistic to believe the gap could be" related to the black puff of smoke seen coming from the lower portion of the right rocket a half a second after liftoff, he said.

Until today, the presidential commission has principally focused on the effect of subfreezing temperatures that purportedly caused the O-ring seals to harden and lose ability to contain the rocket gases.

In dramatic testimony last week, engineers at Morton Thiokol said they recommended against the launch during a five-hour teleconference on the evening of Jan. 27. The engineers said they were concerned about the effect of the cold on the O-rings, but were overruled by Morton Thiokol management under pressure from NASA officials at Marshall.

But Littles said that the O-ring seal may have been damaged seven weeks before the accident when workers were forced to use a hydraulic clamp to bend back into shape one of the right solid rocket segments.

The segments were being stacked here at the Kennedy Space Center, but workers found that the joints failed to match. Technicians then used the clamp to "round out" one of the segments.

This in turn could have caused a "scraping" of the O-ring rubber, he said.

"That does lead to the potential of having an O-ring damaged as you're mating it and, of course, we're concerned about that -- that is an active scenario" under investigation, Littles said.

Littles said that on six previous shuttle flights, technicians had used the same hydraulic device to round out rocket segments in order to make the joints fit snugly. But the amount of bending required on Challenger's rocket -- about a quarter-inch -- was the "worst," Littles said.

He said there was no direct evidence proving that the bending contributed to O-ring damage. A Morton Thiokol official in Utah said the operation "was completely within specifications; there was nothing unusual at all about the rounding process."

Littles also said that testing done by Marshall and Morton Thiokol showed that "in static situations," the O-rings will work at temperatures as low as 10 degrees below zero, far below the temperatures on the night before the Challenger launch.

But panel chairman William P. Rogers sharply questioned Littles about the validity of the tests and repeatedly pressed for independent testing.

The testing should be "by an outside group," he said, suggesting that Marshall and Morton Thiokol officials had a self-interest in results that would cast doubt on the cold-weather theory since they had discounted the warnings of company engineers.

"The people who are running these tests are really in a position which . . . if they were successful, they would prove they are right after all," Rogers said.

"It might be nice if there was some outside independent source to work with you people on this," he added.

Rogers also asked that, in view of the number of new possible theories, if NASA's tests could ever conclusively pinpoint one cause of the disaster, thereby allowing the shuttle to fly again.

Also today, the Air Force colonel heading the hunt for the shuttle's wreckage said search teams have identified 17 pieces of debris over a 190-square-nautical-mile area off the Florida coast.

The search team, which includes a flotilla of 11 ships, one submarine and 41 deep-sea divers, has been hampered by stiff winds and strong ocean currents, making it difficult to complete a sonar mapping of the area where shuttle wreckage is believed to be located, Air Force Col. Edward O'Conner said.

O'Conner said about 160 square nautical miles remain to be mapped. More importantly, only two pieces of the crucial right solid rocket booster have been located, he said.

O'Conner said in response to a question from Rogers that it will take "at least a month or two" for the team to locate significant portions of the suspect rocket and as long as three months before that debris can be brought up for investigators.