They also have eliminated the possibility that one leading suspect -- molecular hydrogen -- accounts for any significant portion of dark matter, the composition of which remains a big mystery.

Various lines of evidence have led astronomers to believe that 90 ercent of ordinary matter in the universe, matter made of neutrons and protons, is dark, invisible to them. Now it seems likely, on theoretical grounds, that a significant portion of dark matter may consist of million-degree gas.

The FUSE spectrograph has detected one particular ion, oxygen VI, in new and surprising locations in space. This ion, or charged particle, which also is being studied with the Hubble Space Telescope, can be formed at a few hundred thousand degrees. This result demonstrates the feasibility of FUSE searches for other ions, including neon VIII and sulfur VI, that would indicate quantities of even hotter gas, said Donald York, the Horace Horton Professor in Astronomy & Astrophysics at the University of Chicago and a member of the FUSE science team.

"We have the equipment to test this theory, one way or the other," York said. "The hot gas is the last great hope. If that doesn't turn out right, then we'll have to start over again."

These results, which are based on three months of observations targeting 20 objects outside the galaxy, appear in a special issue of the Astrophysical Journal Letters. The FUSE science team, led by Johns Hopkins University's Warren Moos, includes York at Chicago and other astronomers at Johns Hopkins; University of California, Berkeley; University of Colorado; University of Wisconsin, Madison; Harvard University; NASA's Goddard Space Flight Center, Canada, France, and elsewhere.

"FUSE has its eyes on solving one of the most important riddles in cosmology -- where is the bulk of ordinary matter hiding?" said Michael Turner, the Bruce and Diana Rauner Distinguished Service Professor in Astronomy & Astrophysics at the University of Chicago. "FUSE has great potential to shed light on this important dark matter problem. I am looking forward to FUSE finding the matter that our calculations of nuclear reactions in the very early universe tell us must be there."

The Chandra X-ray Observatory can detect dense concentrations of million-degree gas, but thin concentrations would be difficult to detect with any satellite now flying or likely to fly in the near future except FUSE, York said.

"If you have high density and high temperature, then it's easy to see X-ray-emitting gas. But in the bulk of interstellar space, where most of the gas is, you don't have such high temperatures and high densities," he explained.

FUSE was launched June 24, 1999, by NASA and is funded in cooperation with the Canadian Space Agency and the Centre National d'Etudes Spatiales of France. It carries a spectrograph that York calls one of the finest ever launched. It observes the individual wavelengths of light given off by stars and quasars and reveals the existence of atoms, charged particles and molecules in interstellar gas between Earth and the source. The far ultraviolet region of the spectrum in which FUSE operates is invisible to the human eye, ground-based telescopes and the Hubble Space Telescope.

Until the latest FUSE results, some astronomers had suggested that much of the missing mass was locked up in hydrogen molecules.

"This is such a good spectrograph that many hard problems are trivial to do," York said. One of those problems relates to the amount of molecular hydrogen in the universe.

"That one has been put to rest," York said. "We don't see enough molecular hydrogen to explain the missing mass. It's nice to have that one out of the way."

A tougher measurement that will take another year or two to verify is the ratio of deuterium to hydrogen in the universe. The late University of Chicago astrophysicist David Schramm emphasized the importance of deuterium as a means of measuring the total amount of ordinary matter at a distant and simpler time, when it was all in a soup of neutrons and protons.

York, with colleague Jack Rogerson at Princeton University, made the first galactic measurement of deuterium 25 years ago with a satellite called Copernicus, launched in 1972. FUSE is 10,000 times more sensitive than Copernicus and was built in large part to make deuterium measurements at far greater distances from Earth.

According to prevailing theory, the deuterium-to-hydrogen ratio should be approximately the same throughout the universe. York's original measurements and others that followed have increasingly been interpreted as showing that the ratio varies, at least within the galaxy, because deuterium is destroyed during stellar evolution. But measurements of the primeval deuterium abundance carried out by University of Chicago Research Associate Scott Burles with the Keck telescope in Hawaii indicate a single value.

"That value implies that the total amount of ordinary matter only accounts for about one-seventh of all the matter known to exist," Turner said.

Schramm's last paper, co-authored by Turner, pointed out that the calculations leading to this conclusion could be checked by measurements of the cosmic microwave background, the big bang's afterglow. Recently reported results by two balloon-borne experiments provided the first confirmation of the big-bang calculations of Schramm, Turner and their collaborators indicating that ordinary matter comprises but a small fraction of the total amount of matter in the universe.

Cosmic microwave background measurements being made by John Carlstrom, Professor in Astronomy & Astrophysics at the University of Chicago, with the Degree Angular Scale Interferometer experiment at the South Pole should further refine this test. Turner said he and his team eagerly await the DASI results.

FUSE, meanwhile, will be able to check Burles' deuterium measurements in nearly pristine but closer gas clouds, Turner said. In addition, by measuring the galactic deuterium abundance, FUSE will be able to reconstruct how stars process the primordial gas to make the other elements.

"Deuterium is an ideal tracer," Turner explained. "Stars only destroy deuterium. Thus, less deuterium indicates more stellar processing."