Former government scientists and outside experts appointed by NASA to review the manufacture of the Hubble Space Telescope's mirrors said yesterday that they were often overworked and overwhelmed by the task and did not rigorously check crucial tests done by the mirror manufacturer.

NASA officials said last week that they had found the mirrors to be so badly flawed that a large part of the $2.6 billion project's scientific goals cannot be met until a shuttle flight set for 1993 installs a corrective optical system.

The outside experts, who were members of the Science Working Group that advised the National Aeronautics and Space Administration on building Hubble, said manpower shortages at NASA and high confidence in the mirror manufacturer meant that work was not scrutinized as vigorously as it should have been. The mirrors, among the most sophisticated optics ever attempted, were built by Perkin-Elmer of Danbury, Conn., which has since been renamed Hughes Danbury Optical Systems.

"We could not possibly be the watchdogs," said William Fastie, an optical expert at Johns Hopkins University and one of two specially appointed telescope scientists on the Science Working Group.

In hindsight, the contractor had "more autonomy than they should have had," Fastie said. "{Perkin-Elmer} had a sweetheart deal."

"There was the presumption that these guys all knew what they were doing," said Arthur Davidsen, a Hopkins astronomer and another member of the Science Working Group. Davidsen said that astronomers had confidence in the mirrors because it was widely known that Perkin-Elmer had worked on similar mirrors for spy satellites.

The Hubble has two mirrors that gather starlight and focus it on a tiny spot at the rear of the telescope, where a cluster of cameras and equipment capture the magnified light and send images and data back to Earth. NASA officials say they do not yet know which mirror is flawed, though they have said that the problem appears to be a "manufacturing flaw."

Because of funding constraints a decade ago, sources said NASA set "a manpower ceiling" that hampered review of the Hubble program.

While the mirrors were being fabricated at Perkin-Elmer in the late 1970s and early 1980s, NASA maintained an office at the contractor's facility with a staff of "three, give or take one or two," according to John Humphreys, a subsystems manager for optics at NASA's Marshall Space Flight Center in Huntsville, Ala., who was involved with the project. In 1983, after a major shakeup of the Hubble program, NASA had 20 staff members working at Perkin-Elmer but the mirrors had been completed two years earlier.

"Staffing was perhaps not optimum for the job," Humphreys said. "But it was not all that bad either."

"We were having to oversee a lot more work than a scientist should have on a government contract," said C. Robert O'Dell, formerly NASA's Hubble project scientist and now a professor at Rice University in Houston. O'Dell said the mirror fabrication was scrutinized, but that "does not mean any of us took it on ourselves to examine every detail" of Perkin-Elmer's work.

NASA officials believed that if Perkin-Elmer rigorously tested each mirror individually, they did not need to go to the expense of assembling the telescope and testing it as a unit. That assumption, however, has been questioned since last week's revelation.

Meanwhile, NASA engineers and their contractors continue to puzzle over the source of the hydrogen fuel leaks that grounded the shuttle fleet Friday. NASA technicians last week discovered that Atlantis, scheduled to fly a classified military mission this month, was suffering from a fuel leak similar to the one discovered on Columbia in May.

Engineers have tracked the problem to a pipe that carries liquid hydrogen from the silo-shaped external tank to the orbiter. Attention focused yesterday on the fact that on both Columbia and Atlantis, the alignment of the external tank and the orbiter was off by an eighth of an inch, which could have caused a hydrogen leak, according to NASA associate administrator William Lenoir. More tests are planned this week.