The Fingerprint of Creation

Cosmology rarely makes the front pages. But this past spring, thanks to satellite named COBE, an astrophysicist named George Smoot announced a discovery of truly cosmic proportions.

By Donald Goldsmith|Thursday, October 01, 1992
8:04 A.M., April 23, 1992, Ramada Renaissance Hotel- Techworld, Washington, D.C.: George Smoot, a tall, neatly bearded astrophysicist from the Lawrence Berkeley Laboratory in Berkeley, California, wearing a suit on this exceptional day, stands ready to present the long-awaited results of an experiment aboard the Cosmic Background Explorer, a satellite known to everyone by the acronym COBE. Tomorrow his research team’s discovery will be a feature story in hundreds of newscasts and newspapers, and George Smoot will temporarily become the most famous scientist in the world.

Right now Smoot isn’t worried about the press--he’s worried about the reception he’ll get from his fellow researchers. Several groups have arrived at the meeting to discuss their own COBE experiments, and Smoot isn’t sure their results will support the announcement he is about to make. Why then is he rushing to announce a major discovery that might not pan out? Cold fusion, after all, is still a recent memory. But Smoot is confident his team’s results can stand on their own. He said as much to a group of researchers from MIT only a week and a half before going to Washington. The MIT guys said, ‘We disagree,’ Smoot recalls now. And I still said I’m ready to go. I really am betting my career on this.

Nearly a thousand people--technicians, programmers, engineers, and scientists--have helped make COBE work; some two dozen of them make up Smoot’s team, and each is all too aware that the whole scientific community eagerly awaits their results. Theories that touch on everything from particle physics to cosmology hang on these observations. Smoot’s boss sent him off to Washington with the advice You’d better be right.

The team Smoot heads is responsible for analyzing the data from one of COBE’s three instruments, the DMR (differential microwave radiometers). The DMR has been observing the heavens since November 1989, when COBE was launched into a near-polar orbit, 560 miles above Earth’s surface, to study the cosmic background radiation, the faint microwave glow that’s everywhere in the sky. It is radiation that no star nor any other known object could have produced. This radiation came directly from the act of creation itself--the Big Bang.

According to Big Bang theory, 15 billion years ago all matter in the universe--and all space, too--was crowded together in a state of near- infinite density. The explosion of that inconceivably dense point released a titanic amount of heat into the early universe and lit the cosmos with high-energy radiation. For hundreds of millennia the radiation was so intense that no particles could stay joined together in a single atom. But as the universe expanded, this brilliant glow became a pale shadow of its original self. Expansion diluted its energy, cooling the universe in much the same way that vapor from a spray can cools when released. The original radiation still fills the universe, but it fills a much larger universe than before: each cubic light-year now contains far less energy in the form of this cosmic glow than did a cubic centimeter long ago.

By about 300,000 years after the Big Bang, the expansion of the universe had lowered the radiation’s energy to the point at which it could no longer disrupt atoms. This early time, when matter as we know it began to form, is called the era of decoupling. Ever since that time the original cosmic radiation has traveled freely through space, with its energy constantly decreasing. We now detect it as feeble background radiation that has journeyed across 15 billion light-years, arriving from all directions in apparently equal amounts. During the time it took this radiation to reach us, the radiation that was once here, in the tiny amount of space that expanded to become our turf, sped 15 billion light-years away from us in all directions, becoming grist for a distant civilization’s detectors.

The Big Bang theory explains two key facts discovered earlier in this century: first, that the universe is expanding; and second, that it is filled with this cosmic glow. The first fact came to us courtesy of astronomer Edwin Hubble, who convinced scientists of the universe’s expansion with his careful measurements, beginning in the 1920s, of how rapidly galaxies are flying away from us in every direction. The second fact was discovered by physicists Arno Penzias and Robert Wilson, who in 1964 first observed the cosmic background radiation in the form of microwaves with very low energy, about 3 degrees Celsius above absolute zero, or -454 degrees Fahrenheit.

The smoothness of this radiation--its property of seeming to arrive with precisely the same energy from every direction--testified that the universe was almost completely homogenized when the radiation decoupled from matter. What astronomers were seeing when they studied this glow was a portrait of the universe at the era of decoupling, 15 billion years ago-- apparently a smooth, featureless universe.

But cosmologists who attempted to trace the history of the universe from that time to the present had a problem. The visible universe today, on any scale we have perceived, does not seem to be characterized by homogeneity. It is organized into galaxies, each of which now contains billions of stars. On a larger scale, galaxies cluster together by the thousands. On a still larger scale, clusters of galaxies are gathered into vast superclusters. And on the scale of the cosmos itself, clusters and superclusters of galaxies form a structure that resembles interlocking soap bubbles: enormous, near-empty voids are surrounded by gigantic cosmic conglomerations, of which the largest, dubbed the Great Wall, stretches for 500 million light-years.

How did all this structure-within-structure arise? Cosmologists know that the universe must have achieved its basic organization within a few billion years. In this time the stars formed as subclumps of matter within the galaxies-in-formation. But computer models created to describe the formation of galaxies show that the 15 billion years since the Big Bang is barely enough time to form galaxy clusters through gravitational forces, and not enough time for clusters to clump into larger structures like the Great Wall. Yet gravity seems the only reasonable way to form galaxies, clusters, and walls; no other force can act over such huge distances to produce such giant clumps of matter. According to current theories, structures such as the Great Wall must have grown, through gravitational attraction, out of irregularities in the distribution of matter that existed at the era of decoupling. The more modest the irregularities, the more difficult the formation of these structures must have been.

Thus, ever since the cosmic background radiation was first detected nearly 30 years ago, the search for irregularities has proceeded. Smoot is now about to report to his colleagues on the latest developments in that search. For many weeks rumors have been rife that COBE has indeed found some sort of deviation from smoothness. The DMR team has spent many months analyzing COBE’s first year of observations, pushing their results to the limit of the DMR’s sensitivity. Their goal was to produce a map of irregularities in the radiation--a map that would correspond to the first glimmerings of structure in the universe. Have they discovered something new--or have they simply systematized their errors as a fact that purports to describe the emergence of structure from the fiery void?

The most difficult part of the science team’s work, unheralded yet utterly necessary, has been their attempt to be sure that the DMR’s measurements represent real signals from the universe, not stray noise generated within the satellite or mistakes in the computer programs that analyze the data. The paper describing the search for possible sources of error runs to 56 pages. Even so, Smoot had been troubled by the feeling that the team’s eagerness to announce stunning new results was overwhelming their ability to check the results properly.

What had happened, Smoot recalls, is that we’d been seeing structure all along the way. But we also knew that the systematic errors, the ‘noise’ in the observations, were at about the same level as the temperature variations we were detecting. We’d been working really hard to make a new run through the data with better software and better techniques of analyzing the data, and we were still seeing structure. We checked it and still found a few things wrong, so we were starting a new run. This was sometime in the fall.

Then Ned Wright, a cosmologist at UCLA, saw our first preliminary map. That started stirring up the overall COBE science team. I realized we had to keep focused on making sure it wasn’t an error, that it was really a true signal in the data. And so I said, ‘We should be focusing our attention on how we could have made a mistake. We should reward the person who can find the mistake that we made, so that people are highly motivated to find out it’s wrong, rather than assume it’s right and run on and interpret it.’

Since it was getting close to the holiday season, Smoot offered to buy an airplane ticket to anywhere in the world for the person who could find a mistake in the analysis. Some team members tried, but Smoot’s money is still in his pocket.

Even so, toward the end, Smoot says, people were still not focusing on finding the mistake. Everyone wants to be in on the discovery part, and discovery is like an archeological dig. You’re scraping away layers--you start to expose the pyramid, and it’s getting bigger and bigger. That’s exactly how it was at this time.

8:07 A.M.: Smoot, his habitual mumble well under control, shows a viewgraph (old news to everyone in the room) that summarizes COBE’s first triumph. In 1990 COBE measured the cosmic background radiation’s spectrum-- the amounts of radiation at different wavelengths. The COBE satellite was needed for this measurement because Earth’s atmosphere blocks most of the radiation, leaving only a few wavelengths to penetrate to the surface. Cosmologists had long sought to measure the radiation’s full spectrum to determine whether it matched the spectrum predicted by the Big Bang theory. In January 1990 COBE found that the spectrum matched perfectly, validating the theory to the satisfaction of all but a handful of cosmologists.

COBE’s first triumph raised hopes for a second: the detection of irregularities in the cosmic background radiation, which could show us where galaxies came from. Before the era of decoupling there was no chance for the visible matter we see today to have formed clumps on the scale of stars, galaxies, or clusters of galaxies. Even though random processes would have made some regions a bit more or less dense than others, the background radiation exerted a powerful homogenizing influence: it would have split up any particles trying to come together as atoms, and this constant disruption would have kept any small clumps of matter from gravitating into larger clumps. Before the era of decoupling, you could no more make matter clump together by gravity than you could make a soufflé in a tornado.

Only after about 300,000 years had passed, and the radiation had ceased to interact with atoms, could gravitational forces between particles make clumps form, persist, and grow still denser. Once a clump had become well established, it would have attracted more and more nearby particles with increasing amounts of gravitational force, as the clumping fed on itself.

What makes the DMR data important is that any lumpiness in the universe existing at the era of decoupling must have left its mark on the cosmic background radiation. The denser regions would have affected the radiation slightly differently than the less-dense regions. Radiation from dense regions would be lower--and those regions therefore colder--since it used more energy to escape the thick clumps. Radiation could more easily shoot through less-dense areas, so they would appear hotter. This fingerprint left on the radiation at the time of decoupling would have endured through 15 billion years, as the radiation moved at the speed of light from all directions and in all directions.

The DMR experiment satellite was designed to find such a fingerprint. Until COBE, all measurements of the cosmic background radiation--from the ground, from balloons, and from rocket-borne detectors- -had registered complete smoothness in the radiation, with no deviations measurable down to an accuracy of one part in 10,000. COBE was built to achieve at least ten times greater accuracy, to better than one part in 100,000. Theorists agreed that if COBE, with its manyfold increase in sensitivity, could not find cosmic irregularities, they must probably abandon their models of galaxy formation based on gravity, a result they dearly hoped to avoid. A negative announcement from the COBE team would be far greater news than the positive result expected by most cosmologists.

At this point in his presentation, Smoot has summarized the DMR’s ability to observe the cosmic background radiation at three different frequencies, measuring only the difference between the amounts of radiation received from two points in the sky 60 degrees apart. Difference measurements are necessary because no direct measurement can achieve anything like the sensitivity of the DMR. (The same is true for many instruments. You can weigh your cat more accurately on a bathroom scale, for example, if you first weigh yourself while holding the cat, then weigh yourself alone, and record the difference as the cat’s weight. If the scale makes consistently high or low errors at a given time, those errors cancel when you make the subtraction.)

The DMR needs six months to complete a map of the sky in all directions, taking care never to point toward Earth or the sun. To map the entire sky, COBE sacrifices some resolving power--the ability to see fine detail--in favor of a global view. Each of COBE’s individual observations produces a pixel, or picture element, that spans roughly 2.6 by 2.6 degrees, 27 times the area covered by the full moon. To make an accurate map from difference measurements, each of the 6,144 pixels that cover the entire sky must be observed thousands of times.

In its first year of operation, the DMR experiment collected some 380 million pixels. Each area of the sky was observed in two entirely separate instrument channels (to check on the instrument noise) for each of three different wavelengths. Finally, confident that the DMR had found irregularities, the science team proceeded to the heart of their project-- analyzing the pixels to see whether they could find any pattern in these deviations from smoothness.

This represented a monumental task. Two important patterns clearly exist, but they do not correspond to any structure in the early universe and must be subtracted from the observations before the search for cosmic irregularities can proceed.

One pattern is the emission of radiation from the Milky Way, our own galaxy, which adds to, and interferes with, the radiation from much farther away. The second pattern arises from the cosmologically modest motions of the solar system and the Milky Way in space. These motions produce a characteristic pattern, called a dipole, in the cosmic background radiation: the radiation appears a bit stronger in the direction toward which the Milky Way is moving, and a bit weaker in the opposite direction. Fifteen years ago Smoot headed the team that found this effect. Now his much larger team has had to remeasure the effect, with much greater sensitivity, in order to proceed with the search for any irregularity dating back to the early universe.

8:08 A.M.: Smoot, at the peak moment of his scientific career, flashes a grin of triumph as he describes what his science team has found: We have a quadrupole, he says, in the language used by those who search for deviations from smoothness. What he means is that they’ve found higher than average radiation emanating from two different directions in the sky, and lower than average radiation coming from the two opposite directions. In fact, there are differences along many different directions--the quadrupole is just part of the full description of irregularities, which involves still more complex items with names such as octupole and hexadecupole. These irregularities, says Smoot, unlike the dipole, arise not from our motion in space but from the cosmos itself. They are the oldest and largest known structures in the universe. The smallest of them now spans a region larger than the Great Wall.

What COBE has found are not structures in the traditional sense. Instead, they are regions whose temperatures at the era of decoupling differed by about one part in 100,000 from the average. These differences are measured in microkelvins, or millionths of a degree Celsius: a typical deviation is a mere 16 microkelvins. Nevertheless, these deviations from smoothness are immensely significant. They provide crucial support for all theories that explain the formation of galaxy clusters by gravity. And since astronomers have no other viable theories, their relief is palpable, summed up in the reaction that one of the leading theoreticians, J. Richard Bond of the Canadian Institute for Theoretical Astrophysics, has for the newsreporters: It’s about time.

8:11 A.M.: Smoot’s scientific talk is over, to be followed by presentations from other researchers on the COBE science team. I had to wait to see what they were going to say, Smoot says, so there was a tense part. When they displayed their own temperature findings, he breathed easier. I looked at that curve and saw an amplitude of 14 microkelvins, which is close to 16. So there’s no way they could say we were wrong.

The presentations also include an attempt, by Ned Wright, to fit the results into the welter of cosmological models that have been developed since the cosmic background radiation was first detected. Wright casually shows a viewgraph that threatens to discard years of work by several dozen theorists; some of them are in the room to nod agreement.

The sweetest part of Smoot’s talk lay in its last minute, when he presented a viewgraph that described the irregularities in the cosmic background radiation in more detail. The irregularities carry a crucial message from 15 billion years ago that goes well beyond the key fact that they exist.

The message lies in the relative numbers of irregularities with different sizes. The size distribution is important because the numbers of different-size irregularities reveal which types of cosmological models can furnish a viable explanation of galaxy formation. For example, the DMR results seem to rule out an entire class of models based on cosmic defects. A cosmic defect is a region of immensely high density that might have been one of the seeds of galaxy formation through gravity. Cosmic-defect models predict far more small-scale than large-scale irregularities. But COBE found essentially the same number of irregularities of all sizes within the range that it measured. For now, cosmic-defect models seem ready for the discard pile of ideas that look fine on paper but do not describe the actual universe.

If cosmic defects are out, what model is in? The size distribution of irregularities revealed by the DMR observations supports the inflationary model of the universe. Inflation adds a new layer to the standard Big Bang theory and gives a description of the universe in its earliest fraction of a second. According to inflation, everything that we call the universe began at the Big Bang as a submicroscopic patch of space- time, pinched off from a much larger metauniverse. The patch began to expand much more rapidly than the speed of light, and in a cosmic eye blink became far larger than the universe we can see today.

When the inflationary model was created during the early 1980s, few noticed that it made a prediction about any irregularities that arose in the primordial universe. According to inflation, we should find irregularities of all different sizes in about the same numbers--just as the DMR experiment observed.

This prediction was a by-product of the basic inflationary model, made at a time when no one expected irregularities to be found soon. For this reason, the DMR experiment’s findings were far more impressive than they would have been if the inflationary model had been conceived to explain the observations after they were made. As University of Pennsylvania physicist Paul Steinhardt, one of the theory’s originators, puts it, It’s a tremendous extrapolation to suggest from what we observe today that the entire universe inflated from a tiny patch of space-time. The DMR results, Steinhardt says, lend new support to the bold model.

Before the announcement of the DMR results, attention had focused on another prediction of the inflationary model, which seemed more directly testable. Inflation predicts that the universe contains a hundred times more dark matter than the matter that shines in stars.

Dark matter is a hot topic in astronomy, not only because of the inflationary model. Dark matter is matter that does not shine, matter whose existence we deduce from the gravitational force it exerts on visible matter. Astronomers can determine the amounts of these gravitational forces by observing the motions of stars and galaxies: greater amounts of matter produce more rapid motions. For example, the stars in the outer regions of our own Milky Way, and of other similar galaxies, will move more rapidly if the galaxy includes more mass. From counting stars and galaxies, astronomers can estimate the mass of visible matter. By measuring the motions of stars, they can determine the total mass of matter--both visible and dark--that should be present.

During the past two decades astronomers have concluded that giant galaxies like our own contain at least ten times more dark matter than visible matter. No one knows what the dark matter is, but many theorists believe that dark matter is highly unlikely to be anything like the matter we know--not rocks nor burned-out stars nor cosmic specks of dust. If this is true, we might do better to rename the dark matter transparent matter, because it seems likely that it does not interact at all with our familiar universe--neither with ordinary matter nor with radiation like that observed by COBE.

For cosmology, the most fundamental unanswered question about dark matter is, How much of it exists? If the universe contains only ten times more dark matter than visible matter, then according to all Big Bang models the universe will expand forever, unable to pull itself together by gravity. But if the universe contains 100 times more dark matter than visible matter, then the larger amount of dark matter will make the universe eventually cease expanding and start contracting.

The inflationary model specifically predicts that the density of matter in the universe exactly equals the critical value that divides models with eternal expansion from those with eventual contraction. Therefore a measurement of the actual density would allow us to determine whether or not the inflationary model is valid--and whether or not the universe will expand forever.

It is believed that the universe contains far more dark matter than visible matter. But precisely because the dark matter is invisible, our prospects for making a direct and accurate determination of the total density of matter are poor. So how can we hope to resolve the mystery of the amount of dark matter in the universe? The answer lies in the DMR results from COBE.

The best way to determine how much dark matter the universe contains consists of finding ways to test the various models of how galaxies form. In all these models, dark matter does two marvelous things to help make galaxies grow from less-dense clumps of matter. First, because the universe contains so much more dark matter than visible matter, any theory that relies on gravity to make clumps form relies automatically mainly on dark matter. Second, because dark matter does not interact with light or any other form of electromagnetic radiation, it would have experienced none of the homogenizing effect that the cosmic background radiation exerted on ordinary matter in the early universe. As a result, the dark matter could have begun to clump together long before the era of decoupling, 300,000 years after the Big Bang, when the background radiation ceased to interact with ordinary matter.

In combination, these two facts about dark matter--the gravitational force that it provides and its head start in forming clumps-- give theorists a fighting chance. Using the models of dark matter most favorable to forming clumps, they can now--though just barely--make galaxy clusters form from the types of irregularities that the DMR experiment detected. But are the models right?

In seeking to test the inflationary model, or any other model purporting to describe the early universe, cosmologists now have two needs. They need data that will either confirm or disprove the irregularities already observed by the DMR experiment. In addition, they need to find and measure irregularities in the cosmic background radiation at smaller scales than the DMR can study--scales that correspond to the formation of galaxies and galaxy clusters. Fortunately, cosmologists should not have to wait long, because a nicely interlocking combination of experimental evidence should emerge within the next year or two.

To begin with, there’s still COBE. The satellite continues to perform nobly, producing a steady stream of data from the DMR experiment for eventual analysis. But in addition, there are now in progress three other types of experiments to observe the cosmic background radiation. First, a series of balloon flights will carry detectors 100,000 to 120,000 feet above Earth, altitudes to which most of the cosmic background radiation can penetrate. Second, observations are being made from the South Pole, where the air is remarkably free of the water vapor that absorbs most of the radiation. Third, radio observatories such as the Owens Valley Radio Observatory in California are being fine-tuned to sensitivities well beyond their original capabilities. All three types of experiments have already gathered data, and all received a great psychological impetus from the DMR results announced in April.

The COBE results have breathed new life into what had threatened to be a lifelong search for cosmic irregularities. Theorists are now geared up to perform the herculean tasks of investigating which models are able to fit the DMR experiment and the data to come from other observations. It’s going to be a really hectic next year or two, says Steinhardt.

All in all, COBE has been one marvelous satellite. It verified that the cosmic background radiation did indeed arise in a Big Bang universe, for the spectrum of the radiation exactly matches the prediction of the Big Bang model. The DMR experiment also found the irregularities that signal the first phases of clumping in the cosmos. The DMR data may yet reveal the amount of dark matter in the universe and will thus test whether the inflationary variant of the Big Bang model is correct. Another COBE instrument, the diffuse infrared background experiment, is mapping radiation emitted by dust particles in the Milky Way with a clarity never before attained. Not bad for a satellite that had to be quickly redesigned to a smaller size and mass after the Challenger accident in 1986 essentially eliminated plans to launch COBE from the shuttle.

Noon: The COBE researchers, who have known all along how important their findings were for their colleagues, discover as they face the press that the world has seized on those findings as a crucial step in our understanding of the cosmos. I went up to get a snack, says Smoot. I’d had breakfast at five in the morning, and a couple of viewgraphs didn’t arrive till the time of the press conference, so I was trying to get composed. Then I walked in that room. That’s when I realized it was a big story. I thought, well, if we really do a good job we’ll get on the front pages of the newspapers. Then the guy comes up to me and says you can’t use the microphone, you’ve got to get up and talk at the podium. I thought, oh my God, and I walked up on the stage, and the lights came on, and it was like looking at the sun.

That’s when it sort of hit me, what it really meant. And so I can’t remember all that I said at the press conference, but apparently I said, ‘If you’re religious, it’s like looking at God.’ Looking back, it was the right kind of thing, but when I first heard it, I was horrified. In fact, if you look at what most cosmologists were saying, about half or two- thirds were in some sense mystical or religious: it’s the Holy Grail of cosmology, it’s the birth of the universe, the handwriting of God. It’s the natural description.

Fond of visiting archeological sites, Smoot compares his team’s discovery to the uncovering of an ancient pyramid: the real credit goes to those who engaged in the relatively boring labor of careful excavation. I would be terribly embarrassed by the publicity if it weren’t for the fact that I thought I was doing some good. Here’s a story about science, about good science, that doesn’t offend religion, that doesn’t offend a lot of things, that’s not going to be used to trash the environment. I already heard from one of my cousins that her kids want to be scientists now, because you get to be famous and you get to discover the universe.

After the press conference Smoot goes to the pressroom, where he continues to answer questions. It went on for hours, he says, and I was there, giving talks and interviews. Finally, around five o’clock, I said to Phillip Schewe, the press officer for the American Physical Society, ‘Have you ever seen anything like this?’ And he said, ‘Not since cold fusion.’

A Brief History of the Universe

The Big Bang: Some 15 billion years ago, everything in the known universe was crowded into an infinitely dense and hot point that violently exploded. At that time the four forces of nature that govern the universe today--gravity, electromagnetism (which sparks lightning bolts and directs compass needles), and the strong and weak forces which hold atoms intact-- were welded together by the unimaginable heat into a single unified force.

By 10-43 second the universe had cooled to 100 million trillion trillion degrees. Gravity became a distinct force, but matter remained an indistinguishable soup of collisions more energetic than anything witnessed in the universe today.

After a mere 10-34 second the temperature had dropped to a billion billion billion degrees--cool enough for the first wisps of matter to coalesce. Quarks (black), the building blocks of protons and neutrons, emerged, as did electrons (pink) and similar indivisible particles. As the first pieces of matter formed, the charged conditions also created their antimatter counterparts. At this point the strong force, which holds protons and neutrons together, split off. Three distinct forces--gravity, the strong force, and the electroweak force--began pulling and pushing at the primordial bits of matter.

By 10-10 second the electromagnetic and weak forces had separated. From that moment on, the four forces began shaping the universe.

At 10-5 second the universe had cooled to a trillion degrees. When quarks crashed into one another now, they stuck together. At the lower temperatures, particles that rammed into the newly formed protons (blue) and neutrons (green) could no longer knock them apart. Antiquarks also came together and formed antiprotons. For every 10 billion particles of antimatter, the universe contained 10 billion and 1 bits of matter. In the still-dense broth of the early universe, flecks of anti-matter and matter collided, annihilating each other in a flash of particles of light called photons (purple). By now the energy level of the universe was too low to build new quarks and antiquarks, so virtually all antimatter was destroyed, leaving only a fraction of the original mass of the universe. The resulting photons ricocheted around, unable to escape the dense environment.

One second after the Big Bang, electrons and their antimatter counterparts, positrons, similarly annihilated each other; because electrons slightly outnumbered positrons, only electrons were left.

One minute after creation, the now relatively sluggish neutrons and protons banged into each other, stuck together, and formed the nuclei of helium, lithium, and heavy forms of hydrogen. After this brief instant in time, the temperature fell below a billion degrees and the density of the ever-expanding universe was too low for such nuclei-forming collisions to occur again. (Billions of years would pass before stars forged helium into heavier elements such as carbon and oxygen--the building blocks of life.)

When the universe was 300,000 years old and 3,000 degrees hot, the assembled nuclei captured electrons as they whizzed past and formed the first atoms. As electrons were constrained to orbiting atoms, the photons that the electrons had once interacted with began to streak the universe with light. Because the atoms were unevenly scattered throughout the expanding cosmos, the photons emerged in a rather spotty pattern seen today as slight variations in the microwave background radiation--the afterglow of the decoupling of energy from matter.

After one billion years or so, the pull of gravity had caused atoms to coalesce into clouds of gas. Galaxies formed as the nascent clouds continued to swirl together, perhaps around bits of very dense but undetectable dark matter. In another billion years the galaxies themselves began to group together into superclusters and gigantic structures--bubbles of space with walls of thousands of galaxies surrounding voids millions of light-years across.

After 3 billion years the stars began to shine. Earth and our solar system formed 10.4 billion years after the Big Bang. After basking for another 2 billion years in the light of our nearest star, our planet sprouted its first traces of life.

Today perhaps 100 million species of plants and animals inhabit Earth.
Comment on this article

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar

Log in to your account

Email address:
Remember me
Forgot your password?
No problem. Click here to have it emailed to you.

Not registered yet?

Register now for FREE. It takes only a few seconds to complete. Register now »