The US national laboratories, ranging from small environmental stations to the Fermilab National Laboratory's giant atom-smashing facility, have often been called the ''jewels'' of American research and development.
But Reagan administration officials have long complained that these jewels don't shine quite as brightly as they used to. And the federal government is now making plans it hopes will bring the national R&D effort back to its former brilliance - at least in some areas.
''It is well past time to take a fresh look at this situation,'' said Dr. George Keyworth, White House science adviser, when he released a study of the federal labs last week.
If this effort has a main theme, it is to concentrate resources on the most promising projects.
The government, for instance, is taking a new look at the construction of various accelerator labs - the giant sites of nuclear research, where bits of matter are smashed at terrific speeds in order to glean insights into nature's basic forces.
On one hand, Energy Secretary Donald P. Hodel announced Wednesday that he will recommend construction of a $147 million National Electron Accelerator Lab in Newport News, Va. The announcement ended a fierce tug of war between Virginia and Illinois over the project. (To win, Virginia offered such sweeteners as 35 new faculty positions in nuclear physics at nearby universities, a move that would increase professors in the field by 10 percent.)
On the other hand, a panel of physicists has recommended halting construction on a high-energy physics accelerator at the Brookhaven National Laboratory on Long Island, after four years and $200 million worth of work.
This project - now known as the Colliding Beam Accelerator - has been hit by rising costs and design delays. If completed, the facility still wouldn't be as powerful as similar European machines.
These moves are taken against the background of last week's release of a federal lab study by the White House Science Council.
The council, an advisory panel of non-government scientists, cited ''serious deficiencies'' in the federal lab system.
Some labs, complained the council, don't have clearly defined missions for much of their work. The panel particularly criticized Department of Energy labs that veered off into research on alternate energy sources in the mid-'70s. It pointed to such single-minded facilities as Fermilab, in Illinois, as good examples.
The Science Council also said the uncertainty of the appropriations process hurts the labs, as scientists waste time in budget meetings and projects get delayed.
Last year's budget crunch in Congress, for instance, has slowed down work this year at Fermilab.
''We didn't know our budget for the fiscal year until after three months of the fiscal year had passed,'' says Leon Lederman, Fermilab's director. As a result, the $130 million upgrading of Tevatron, Fermilab's particle-smashing accelerator, has been delayed.
Multiyear budgeting, with Congress approving funds for longer than a one-year period, ''would be the best way out,'' says Dr. Lederman.
Overall, he agrees with most of the Science Council's findings. ''They say Fermi's a good lab,'' he laughs. ''How can I argue?''
The Science Council also said that many labs can't keep good scientists, because they can't match private-sector pay scales. Government-owned and -operated labs, said the council, should be able to bypass civil service salary caps and offer their star scientists more money.
The panel also concluded that the US ''can no longer afford the luxury of isolating its government laboratories from university and industry labs.'' Even the giant Defense and Energy Department labs, said the panel, should be able to cooperate on projects with the private sector.