Twilight falls, the sun sets, and on a clear night the sky metamorphoses into a pointillist canvas whose pastels have been replaced by twinkling gobs of titanium white. Pretty enough-unless you know what you're missing: more and more, astronomers are coming to realize that stars are but a tame sideshow compared with the rest of what's out there. If our retinas registered wavelengths shorter than the color of a deep violet pansy and longer than the red of a ruby, we would witness a firmament of such turbulence and incandescent splendor that it would make Fourth of July fireworks look like a backyard sparkler. We would see black holes slurp up their neighbors and jets of plasma stream across the sky with the energy of 100 million suns. We would see stars-so small that they would fit comfortably in Lake Tahoe-spin an astounding 643 times per second. We would see neutron stars so dense that a teaspoonful of their tiny bodies weighs 1 billion tons, and clouds bigger than our solar system collapse like a poorly timed souffle.

But the cones and rods of our eyes do not resonate to the cosmic. To find the fugitives hiding in the electromagnetic spectrum requires new instruments-the kind that will debut in the 1990s. "The greatest astronomical discoveries have always been made by better technology, rather than by the brightest minds, " says astronomer Stephen Maran of NASA's Goddard Space Flight Center. No genius suspected that a massive black object-larger than any theory can account for-may have been lurking in a galaxy only 300 million light-years away from us. Yet in April three astronomers using an optical telescope in Hawaii announced they'd found one. No theorist imagined that galaxies form clumps a regular 400 million light-years apart. Yet there they were, spied by optical telescopes in Arizona and Australia last year.

Telescopes of the 1990s promise more surprises. The largest in the world-the $94.2 million Keck, in Hawaii-will come on line by the end of 1991, bringing the universe into focus as it captures more of the spectrum from more of space and time. A $93.3 million Keck twin, also on Mauna Kea, will become operational by 1996. Advanced optics perfected in spy satellites and lasers, which the Pentagon declassified last week, will make infrared telescopes thousands of times more powerful. They'll pick up heat from stars so distant that by the time the radiation reaches Earth it wouldn't warm an amoeba.

New radio telescopes will capture millimeter radiation, emissaries from galaxies whose stars are literally bursting into existence. A hypersensitive optical telescope planned for Apache Point, N.M., will, by the turn of the century, enlist digital computers to map 100 times more galaxies than cosmic cartographers have ever recorded. And just as the curtain falls on the millennium, the European-built (and aptly named) Very Large Telescope will begin eying the heavens from a mountaintop in Chile: its four 8-meter mirrors will make it the largest in the world.

Other instruments will sally forth to meet incoming photons-particles of electromagnetic radiation-before they are blurred, smeared and absorbed by the Earth's atmosphere. Despite the woes of the Hubble Space Telescope, NASA has not given up on space-based astronomy. The agency lofted the $600 million Gamma Ray Observatory from the space shuttle in April. From 280 miles up, the GRO's four large instruments capture powerful radiation such as those emitted by "gamma-ray bursters," which suddenly go off like a flashbulb. This week at the annual meeting of the American Astronomical Society, Gerald Fishman of NASA presents the first data from the GRO, which suggest that the bursts emanate from both within and beyond our Milky Way, perhaps in thermonuclear flashes incited when a comet, gas cloud or another star smashes into a neutron star. The $1.6 billion Advanced X-ray Astrophysics Facility (AXAF) may et a ticket for a 1997 shuttle; it should glimpse emissions associated with black holes. A $1.3 billion Space Infrared Telescope Facility (SIRTF), which should see stars forming and galaxies aborning, may get congressional funding by 1994 for launch soon after 1999. "The 1990s," declares physicist John Bahcall of the Institute for Advanced Study in Princeton, N.J., "Will be chockfull of astronomical discoveries."

The instruments of the 1990s are intended to solve cosmic puzzles left over from the 1980s. In that decade, a burst of discoveries radically changed our notions of the universe. Infrared and optical telescopes in orbit and on the ground found galaxies arrayed in a complex, structured way-what astrophysicist Alan Dressler of the Carnegie Institution's observatories in Pasadena, Calif., sees as "more like sculpture than spatter." The galaxies sparkle like glitter on soap bubbles: galaxy upon brilliant galaxy lie on thin, curving sheets surrounding immense voids. This finding challenged the dogma that the universe is homogeneous, without structure. What wasn't explainable was simply humbling. X-ray detectors gathered evidence that humongous radiation-spewing beasts lurk in the centers of violent galaxies and left theorists scrambling to find a power source that could explain them. Optical observations of stars in galaxies implied that some 90 percent of the matter in the universe is zipping around totally undetected by any instrument, let alone the human eye. Somewhere out there is a shadow of creation, an entire unseen universe.

Finding it involves time travel on the grandest scale. Because the particles that telescopes catch began their space odysseys eons ago, says John Huchra of the HarvardSmithsonian Center for Astrophysics, "we are seeing things as they were when the light was emitted." A telescope that snares some of the trillions of photons raining down from the cosmos is collecting messengers of creation, bearing clues to how the universe evolved. "With today's instruments we can see 7 billion years back," says astronomer Sandra Faber of the University of California, Santa Cruz. "What we would like is to see the universe when it was 1 billion to 2 billion years old," some 12 billion years ago.

When it is completed late this year, the Keck Telescope will do just that. Perched 13,000 feet above the surf, on a wind-swept and desolate ridge on Mauna Kea, it will be the largest telescope in the world. In astronomy, bigger is better. The larger its mirror, the more photons a telescope gathers, and thus the more clues to the cosmic past it snares. But the Keck's power comes as much from brains as from brawn. Big mirrors sag under their own weight, distorting the images they capture. To get around this limitation, the Keck has 36 six-foot-wide hexagonal mirrors, weighing 880 pounds apiece. Twice a second, 168 sensors and 108 "actuators" electronically guide and align the mirrors to an accuracy of one millionth of an inch (diagram), so that they form a perfectly smooth, parabolic 10-meter mirror. Last Nov. 24, with nine mirrors in place, Keck scientists aimed it at a spiral galaxy called NGC 1232, 65 million light years (390 quintillion miles) away. The light Keck collected began its journey about when the last dinosaurs thundered over the Earth. Despite cloudy weather, the image was as good as the best from the 200-inch telescope at Palomar Mountain, Calif. With all 36 mirrors in place, its collecting power should be four times better.

Among the Keck's first targets will be the enigmatic objects known as quasars. The most distant, brightest objects in the cosmos, quasars emit 1,000 times more energy than 100 billion stars. Quasars' energy is too enormous to come from the nuclear fusion that makes the sun shine, so the Keck, AXAF and SIRTF will all try to glimpse the power source. Most astronomers suspect that quasars are the shining we see when the core of a galaxy contains a black hole, a collapsed star so dense that not even light can wriggle out of its gravitational embrace. As the black hole gobbles up gas and stars, gravitational energy bursts out in the form of radiation, which we see as quasars.

Quasars may also shed their brilliant light on the biggest puzzle in the cosmos: how galaxies, vast collections of hundreds of billions of stars held together by gravity, emerged from the infernal miasma of the big bang. "We hope the new toys will sort out how the universe changed from a smooth gas into a very lumpy thing of stars and galaxies," says astronomer Wallace Sargent of the California Institute of Technology. Most quasars that astronomers see began shining about 3 billion years after the big bang, or 10 billion years ago. Now the astronomical rumor mill is abuzz with talk of the discovery, from Palomar, of a quasar that is older and brighter than any previously detected. Its spectrum indicates that it formed when the cosmos was a mere 6.7 percent of its current age or during the first 1 billion years. This ancient floodlight can claim with only slight exaggeration to have been present at the creation, and thus is a clue to the universe's youth. "If we can find out when quasars got switched on, that would tell us when galaxies started," says Sargent.

The theory of galaxy birth is straightforward. Chance fluctuations soon after the big bang created tiny clumps of gas. These denser regions had more gravity than sparser ones, and so attracted matter, leaving sparse regions emptier. "The rich got richer and the poor got poorer," says Faber.

But the theory has a flaw. Rich regions may well get richer, but recent observations have glimpsed structures so huge that they would have required outlandishly large nest eggs. In 1986, Dressler, Faber and five colleagues found that the Milky Way and its neighbors are rushing headlong toward the Virgo constellation, apparently pulled by a powerful gravitational siren 50,000 quadrillion times the mass of the sun. This "Great Attractor" is larger than cosmic structures have any right to be. In 1989, Harvard's Huchra and Margaret Geller unveiled an extraordinary map of a bow-tie-shaped 1 percent of the cosmos. In it was a "great wall" of 1,700 galaxies 500 million light-years long-another structure more immense than anyone thought likely. This January, astronomers led by Will Saunders of Oxford University reported that 2,163 galaxies, covering 74 percent of the sky, formed superclusters hundreds of millions of light-years across.

Even more bizarre is the discovery last year by A. S. Szalay of Johns Hopkins University and colleagues that galaxies clump together every 400 million light-years. Says Edmond Bertschinger of the Massachusetts Institute of Technology: "The periodic structures are like periodic extinctions for geologists. Why should the process that made galaxies pick out that pattern? It is so beyond our understanding that theorists dismiss them for the time being"-hoping they're an illusion.

That won't do for long, however. To break the logjam, NASA dipatched a little robotic emissary to find evidence of the seeds from which galaxies in all their wild permutations grew. Called COBE, the Cosmic Background Explorer, it was launched in November 1989 to map the barely perceptible energy-100 million times fainter than that emitted by a birthday candle left over from about 100,000 to 300,000 years after the big bang. Light waves that began their journey then as high-energy gamma rays have been stretched out by the expansion of the universe and now appear as weak microwaves. COBE finds that the radiation is uniform to one part in 25,000. Put another way, the temperature of the universe-2.735 degrees above absolute zero-hardly varies from one side of the sky to the other.

This uniformity implies that the early universe was extraordinarily homogeneous. And that makes galaxy theorists uneasy. Where are the lumps and bumps that served as the seeds of the Great Wall, of the Great Attractor? "The measurements haven't reached the point where we should start sweating," says Dressler. But soon they might. COBE is looking for lumps as small as one part in 100,000. If it finds none, astronomers will be hard pressed to explain how great structures formed.

Just as some structures are too large to fit comfortably in the standard theory, so some are too old. Take the bizarre creatures called jets, beams of charged particles traveling at more than 1,000 miles per second that extend out trillions of miles. They are "the largest single objects in the universe," says David De Young of Kitt Peak National Observatory. And they carry more energy than all the stars in the Milky Way produce in 100 million years. A jet probably derives its power from a massive spinning black hole that eats its surrounding disc of rotating gas. When magnetic fields speed up and focus some gas before it is consumed, a jet is born. This takes time. Since the oldest jets come from galaxies nearly as old as the universe, standard theories "may not give us enough time to produce them," says De Young.

In science, crisis often precedes triumph. When medieval astronomers tried to patch up the Ptolemaic picture of the solar system with a confounding mess of orbits within orbits, they paved the way for the Copernican revolution. Astronomy in the 1990s is at just such a crossroads. As its pet theory of cosmic evolution teeters, a riotlof new ideas are bursting into journals.

One hypothesis raises the unhappy possibility that something might have erased the galactic seeds that telescopes seek. "Suppose there were seeds, but the evidence for them has been wiped out," says John Mather, COBE's chief project scientist. That could happen if the particles that seeded galaxies later decayed, or if the cosmos somehow became opaque to primordial radiation. In that case, we might never see where the Milky Way and the other galaxies came from. Alternatively, the rich-get-richer picture may be all wrong. Theorist Craig Hogan of the University of Washington thinks that galaxies might have been midwifed by powerful cosmic brooms formed of radiation. In the early cosmos, he says in a just-published theory, X-rays or ultraviolet rays emitted by the decay of subatomic particles called massive neutrinos might have pushed primeval atoms faster and faster. (Something similar can be done in the lab.) The rays would sweep up far-flung bits of matter. "If the radiation is sufficiently intense," says Hogan, "the [result is] perturbations of the type required to generate galaxies and large-scale structure." If COBE found radiation with a special signature, it would provide strong support for this theory.

In "Life on the Mississippi," Mark Twain described science as an endeavor in which "one gets such wholesome returns of conjecture out of such a trifling investment of fact." That's why new instruments are crucial: by glimpsing galaxies being born out of clouds of dust and gas, they would rein in theories long on fancy and short on data. Ground-based eight- and 10-meter telescopes planned for Mauna Kea and South America, as well as the infrared-reading SIRTF, "will locate, identify and study the epoch of the formation of the first galaxies," predicts Frederick Gillett of the National Optical Astronomy Observatories in Tucson, Ariz. (Infrared light carries the story of galaxy formation because light that began its travels so long ago, as visible rays from fired-up newborn stars, has been stretched out by the expansion of the universe. The length of the wave tells scientists, in effect, its age.) The telescopes would be tailored to catch infrared rays and seek galaxies forming nearby, late in the life of the cosmos. If the new telescopes do find that galaxies did not all emerge in a single baby boom, but instead are still forming long after the big bang, astronomers will have to figure out what took them so long.

One key to galaxy formation is literally invisible to all instruments, for it emits no radiation. Studies in the 1980s confirmed that "at least 90 percent of the matter in a galaxy, perhaps as much as 99 percent, is dark," says astronomer Vera Rubin of the Carnegie Institution of Washington, D.C. It emits no radiation at all. Astronomers know it's there only through the motions of galaxies and stars: these kinetics can be explained by invoking the pull of some unknown "dark matter" hiding in massive halos surrounding our galaxy and others. What could it be? Candidates include almost everything under, around and beyond the sun: Jupiter-size planets, dead stars, black holes, neutrinos or enigmatic bits of mass, still awaiting discovery, called cold dark matter.

Astronomer J. Anthony Tyson of AT&T Bell Laboratories doesn't know what the dark matter is, either-but he has "seen" it. Using an optical telescope in Chile, he observes arcs of light that form when beams bend around mass. No mass is visible, so Tyson infers that it is dark matter. "For the first time we've taken a 'picture' of the dark matter," he says. Tyson finds that the dark matter is piled up in the center of galactic clusters as if "the luminous and dark matter know about each other. " That suggests that stars and dark stuff have danced together for billions of years-long enough, perhaps, for the dark matter to be the long-sought seeds that grew into galaxies so long ago.

Astronomy is unique among sciences in that experiments are impossible. One can only observe, not manipulate, what nature has dished out. So astronomers settle for second best, making a universe in a grain of silicon. "Astronomy has gone from being a photographic science to being a digital one," says William Press of Harvard. In March, a report by a panel of the National Research Council estimated that nearly 10 percent of working astronomers-about 400-are simulating the cosmos on computers.

One of them, Joshua Barnes of the University of Hawaii, is modeling mergers, acquisitions and extremely hostile takeovers. Some 80 percent of the known galaxies are shaped like discs. The others are fat, blob-like "ellipticals." At least half the ellipticals seem to be youngsters that assumed their present shape long after the big bang. In many, stars orbit every which way, rather than in orderly revolutions, as they do in disc galaxies like our Milky Way. In simulations by Barnes and Lars Hernquist of UC, Santa Cruz, collisions of disc galaxies create exactly the pattern of star orbits seen in ellipticals. Collisions can also create conditions for "star-burst galaxies," which suddenly form stars 100 times faster than the Milky Way does. None of this proves that ellipticals are the wrecks of colliding discs, or that star bursts are triggered by galactic encounters of the close kind. But it's as close as astrophysicists can get to testing their theories until a chance observation wanders into their sights.

That's what happened when three astronomers aimed the 88-inch telescope on Mauna Kea at an infrared galaxy called NGC 6240. One of its two discs of gas, thrown together when galaxies collided as they do in Barnes's computer, seems to spin so quickly (2 million miles an hour) that it requires a huge amount of gravity to hold it together. That gravity may come from a compact object perhaps 100 billion times the mass of our sun, calculated the astronomers in April. Because the mystery core is invisible, the researchers speculate that it may be a supermassive black hole, formed when the colliding galaxies crammed dust, gas and stars into their center. If so, it would be 100 times more massive than is thought possible, says Joss Bland-Hawthorn of Rice University.

Computers can also run the history of the universe. Last year astrophysicist J. Richard Gott of Princeton University produced, with graduate student Changbom Park, the largest-ever cosmic simulation. They start with 4 million particles of cold dark matter, gravity and other laws of physics. They shepherd the particles into lumps consistent with what COBE finds, and let the supercomputer run. The tiny bumps start to grow by attracting surrounding matter when the universe is about 10,000 years old. Eventually they produce quasars as distant as those ever seen, as well as immense voids and galactic clusters, just like Geller and Huchra observe. This result suggests that cold dark matter can create the universe we see today. In his own simulations, MIT's Bertschinger can reproduce the Great Attractor as well as the enormous clusters reported by the Oxford team this January. But he's done it by postulating lumps of cold dark matter within a whisker of being too large to fit the COBE data. If COBE doesn't find lumps soon, "it may be time to run up the white flag on cold dark matter" as the seeds of galaxies, says Bertschinger.

It's been less than 70 years since astronomers discovered that galaxies are not bright clouds in our Milky Way, but island worlds scattered through the cosmos. Having yielded one of her celestial mysteries, nature has managed to stay well ahead of scientists' pursuit of her other ones. In the 1990s, some of the enigmas may finally yield. Yet even as the next generation of telescopes explores the art and architecture of the cosmos, astronomers will inevitably, and serendipitously, stumble upon yet deeper secrets. Vera Rubin is convinced that "some of the most mysterious features of the universe have yet to he discovered." Time and time again, astronomers have been humbled by the realization that nature's imagination is much greater than their own.

All electromagnetic radiation, from X-rays to heat, is composed of particles of light Acalied photons. The rays differ only in frequency. But no single telescope can capture the entire spectrum. So astronomers now use different pieces of equipment in order to view the full panoply of energy emitted by heavenly bodies. The results can be extraordinarily illuminating. Arrayed across this page are the various visions of our Milky Way.

High-energy bursts from neutron starquakes and thermonuclear blasts.

These rays can't penetrate dust and gas, and so show only stars relatively near Earth

This radiation pierces the galaxy's dust, showing the bulge of stars at its center

These heat waves carry images of the galaxy's cool, star-forming regions

Weak emissions (blue) and intense ones (red) mark quasars, pulsars and radio galaxies


When the Keck Telescope in Hawaii is completed this year, it will be the largest in the world. Optical and infrared radiation from the heavens will hit its 36 hexagonal mirrors and bounce up to a detector. In a unique design, the hexagons will be electrically aligned twice a second to an accuracy of one-millionth of an inch, forming a perfectly smooth 10-meter mirror.

1. Each mirror has its own support system, called a whiffletree, as well as sensors along each of its edges.

2. The sensors detect misalignment between adjacent mirrors and update the computer about the mirror's relative positions.

3. The computer directs the actuator to adjust the position of a hexagon up or down, maintaining the perfect surface.

Some targets of today's astronomers sound more like fiction than science. Here's a glossary to help you follow their space odyssey:

Tiniest, densest stars known, made of neutrons. Formed when an old star implodes.

Collapsed star so dense not even light can escape its powerful gravitational field.

Short for quasi-stellar objects. Brightest bodies in the sky; what we see when a black hole swallows gas clouds and stars.

Stretching out of radiation by the expansion of the universe. Amount of red shift reveals distance from Earth to the radiant object.

Invisible particles that may have served as seeds for galaxies long ago.