Regardless of how galaxies populate space or how they emit their radiation, an even more basic question comes to mind: Where did the galaxies come from? How did the grandest of material structures arise from an early Universe comprising a uniform mixture of hot matter and intense radiation? Do galaxies form by engorging already-made stars, or do stars gestate in already-made galaxies? Which came first, stars or galaxies? Not least, how do galaxies evolve, once formed?

Fortunately, we can address these and other questions pertaining to the Matter Era with more assurance than the rather uncertain events of the Radiation Era described in the previous PARTICLE EPOCH. Even here, though, substantial puzzles remain about the details of the galaxy formation process. Astrophysicists are now tackling the issue of galaxy origins and have identified its main problems, but they have not yet solved them.

Lack of good observational knowledge of the galaxies themselves creates the basic enigma. Galaxies can be classified according to their gross morphology and their energy budgets, as done earlier in this GALACTIC EPOCH. But we have thus far no explanation for the observed properties of all the galaxies in terms of, for example, the simple gas laws that describe our rather detailed knowledge of stars, a topic of focus in the next STELLAR EPOCH. Not surprisingly, it’s hard to fathom how galaxies emerge and change when we don’t quite know what they are.

Common Denominators If put under bright lights and interrogated, astronomers would admit they know only a few grand and mutual facts about galaxies. Together, these common denominators are helping us begin to understand the events that produced these most majestic of all objects in the Universe.

First, the galaxies are out there. That’s a telling piece of data, for we do know that the galaxies exist. Our civilization should be mightily proud of that fact; no other life form on Earth knows, or ever has known, of the presence of the galaxies. Yet, their mere existence doesn’t help us much to decipher their origins. Given the galaxies’ expanse and magnificence—in vast numbers, in any direction, as far as our best telescopes can see—we are left perplexed and wondering: Just how did these awesome structures come into being?

Second, there are now no young galaxies. Said another way, no galaxies seem to be originating at the present time. Some may be still growing and developing as they accrete more matter, but none seems to have emerged within the past 10 billion years or so. Since all normal galaxies contain some old stars, and since most active galaxies are far away in space (and thus in time), the bulk of the observed galaxies must have come forth long ago. Whatever the seminal mechanism, it was surely widespread in the early parts of the Matter Era. But if the galaxies originated so prolifically in the younger Universe, then why aren’t they still doing so now?

Yet another common factor derives from the finding that most galaxies house comparable amounts of matter. The capacity of virtually every individual galaxy thus far measured ranges between 10⁹ and 10¹² solar masses. Normal galaxies appear to have about that many stars and, as best can be determined, active galaxies also include roughly this much matter in some form. No known galaxies are much smaller and none much larger. They all seem to average 10¹¹ stars, or their equivalent, much like our own Milky Way Galaxy, give or take a factor of 10 for giant ellipticals or dwarf irregulars. Why should Nature’s grandest intact assemblies have such a narrow range of sizes? What precludes the construction of galaxies containing, for instance, a quadrillion stars?

To summarize, reiterate, and clarify: There is no evidence that galaxies are originating at the present time, nor have any done so within the past many billions of years. Galaxies do seem to be evolving currently, as noted toward the end of this GALACTIC EPOCH, including additional growth as new matter occasionally falls into already established galaxies. But if galaxies were initially emerging only now, astronomers should have spotted some objects having sizes and morphologies somewhere between well-defined galaxies and sheer empty space. We know of no such nearby, amorphous, “half-baked” objects. Furthermore, the regions beyond the galaxy clusters—the intergalactic voids—don’t seem to contain much matter, if any at all. Whenever and however the galaxies did form, they apparently did so very efficiently, sweeping up almost all the (normal) matter available and leaving little behind for further assembly.

What’s more, key theoretical ideas of the next STELLAR EPOCH strongly suggest that stars ought to be forming now within galaxies. The bulk of most galaxies likely formed first, yielding environmental conditions ripe for the later formation of the stars we now see richly populating galaxies. These ideas have been well verified in the past couple of decades by splendid observations in widespread locations throughout our Milky Way, where stars are known to be originating slowly but surely from the galactic hodgepodge of loose gas and dust. Recent stellar census implies that ~10 new stars now form in our Galaxy each year.

Fragmentation To address the issue of galaxy formation, imagine a giant cloud of hydrogen and helium atoms embedded in a weakening sea of radiation, some hundreds of millions of years after the big bang. This giant cloud should not be regarded as filling the entire Universe; rather, think of only a small sector of the cosmos, yet one still millions of light-years across. Figure 2.10 parallels the following description of events, illustrating how clouds might break down into smaller clumps, a process called fragmentation.

FIGURE 2.10 — This artist's conception illustrates how a huge cloud of the early Universe might have fragmented (left to right) to form several galaxies. The scenario shown is actually a top-down approach, which doesn't work well for galaxies, but on much smaller scales works fine for the origin of stars within galaxies.

Although universal expansion continued apace, such a huge clump of mass would not have indefinitely expanded; local gravity slowed the cloud to a maximum size, after which it began to fall back on itself. The cosmic temperature and density had dropped greatly since the onset of the Matter Era. Radiation was no longer sufficiently intense to shatter atomic matter, as fully formed hydrogen and helium atoms were then numerous enough to exert a collective influence of their own. Electromagnetic and nuclear forces bound elementary particles within atoms, while gravity in turn bound the atoms within the giant cloud. All the known forces that now direct the evolution of matter were already operating well enough to grant the cloud some structural integrity of its own. Vast parcels of matter were becoming distinguishable from one another, each isolated in a fragmenting cosmos, a state of affairs strongly contrasting with the uniform mixing and chaotic violence of the earlier Radiation Era.

Despite its growing stability locally and its steady recession globally, the initially homogeneous cloud would have surely experienced occasional fluctuations—small irregularities in the gas density that came and went at random. No cloud, whether a terrestrial fluffy cloud in Earth’s atmosphere, a tenuous interstellar cloud in our Milky Way, or the primordial cloud visualized here in the young and formative Universe, can remain completely homogeneous indefinitely. Eventually, as one atom somewhere in the cloud accidentally moved closer to another, that part of the cloud became just a little denser than the rest. The atoms might have then separated, dispersing this density fluctuation, or they might have acted together to attract a third atom to enhance it. In this way, small pockets of gas arose anywhere in the cloud simply by virtue of random atomic motions. Each pocket of enhanced matter was a temporary condensation in an otherwise rarefied medium. Numerically, such gas fluctuations initially are minute—as little as 0.1% greater than the average density. The whole process is not unlike the billowing clouds of a terrestrial thunderstorm, collecting, growing, and eventually forming condensation nuclei that precipitate rain.

Provided some density fluctuations further developed by gravitationally attracting many more atoms, they could have conceivably grown into clumps of matter that became the seeds of galaxies. Theoretical calculations support the idea that such chancy gas fluctuations could have been the forerunners—protogalaxies—of today’s galaxies. But—and this is an important but—these same calculations suggest that, given the slow rate of chance encounters, the protogalaxies would only just be forming at the present time. Yet, as noted, astronomers have no evidence whatever that galaxies are now orginating; we have found few, if any regions caught in the act midway between full-fledged galaxies and intergalactic nothingness.

Extremely long times—typically several tens of billions of years—are needed for enough randomly moving atoms to coalesce into a large pocket of gas that can be rightfully called a galaxy. This lengthy duration is not surprising given the absolutely gargantuan quantity of atoms in a typical galaxy—namely, ~10⁶⁸ atoms. How do we know that? Well, each galaxy houses ~10¹¹ stars, each star averages 10³³ grams, and each gram has ~10²⁴ atoms, all of which adds up to a very big number. That’s why astronomers prefer scientific notation, in which case the number is indeed 10⁶⁸ atoms—clearly an awful lot of atoms to collect regardless of the notation used. Consequently, it takes a great while for Nature to do it at random.

But—and this is an even bigger but—no scientist ever said that galaxies were built by random events, by chance and chance alone. Some philosophers of science, or historians of science, or others who, like postmodernists, tend to criticize the methodology of science yet have never practiced science themselves, have occasionally made such claims to champion the cause of pure chance. By contrast, few natural scientists have ever argued that chance and only chance plays a role in any physical phenomenon. Rather, Nature almost surely operates by combining chance with necessity, randomness with determinism—a basic, unifying issue to which we shall return several times in this Web site, especially when describing the origin and evolution of life in later epochs.

A time of several tens of billions of years is of course well longer than the current age of the Universe—meaning that no galaxies should now exist if chance prevails. So, despite the likelihood that random density fluctuations in an otherwise homogeneous gas could have eventually produced galaxies, it’s unlikely that the galaxies we now see emerged strictly in this way. Chance cannot be the sole factor governing the origin of these immense cosmic systems. Still, the idea of naturally arising gas inhomogeneities remains a powerful concept, for it’s a reasonably well understood process not requiring any unknown forces or unique conditions. If we could find an agent or mechanism, or some means or another, that would accelerate the growth of the gargantuan number of atoms needed to form a galaxy, then we might begin to understand their origins.

To clarify the oft-misunderstood issue of chance versus necessity: Chance surely does play some role in galaxy formation, especially as the initial trigger that starts the fragmentation of primordial clouds. Other, more deterministic agents in the early Matter Era, such as turbulence or shocks, could have enhanced the growth of the inhomogeneities, permitting myriad galaxies to form within a timescale shorter than the age of the Universe. Or perhaps the elusive dark matter helped to accelerate the emergence of galaxies as noted below. Or maybe the seeds of the galaxies were sown at the quantum level much earlier, in the chaotic events of the Radiation Era as also noted below. Whatever it was and however it worked, the enhancement process must have been surprisingly effective since observations clearly imply that the bulk of virtually all galaxies formed long ago, apparently within the first few billion years after the big bang.

The problem of galaxy formation is currently a tough one for astrophysicists. Its solution has exasperated many brilliant minds and is still not yet in hand. The origin of galaxies appeals to theorists having fertile imaginations (to visualize conditions so long ago) and computing skills (to keep track of all those atoms), and especially to those willing to make unorthodox assumptions. This is one of the trickiest areas of astronomy to appreciate, for few hard facts are known about galaxies, and fewer still about the physical events that formed them long ago.

One hard fact that is clearly known, however, is the first one noted above: Galaxies do exist. They populate the Universe in great abundance. Somehow they came into being, and somehow they got to be where and when we find them in space and time. Let’s consider in greater detail some of the specific galaxy formation scenarios recently proposed by theoreticians.

Turbulence Astrophysicists today seek to identify ways that random gas fluctuations might have been enhanced earlier in the Universe. After all, well-founded physics roundly suggests that these minute density fluctuations must have indeed arisen, yet their growth would have been too slow to fabricate the galaxies now seen. If some factor could be found to speed the growth of the density irregularities, the problem might be solvable. One such possibility is to assume that the Universe was quite turbulent long ago—a not altogether unreasonable idea since turbulence involves the inevitable “confusion” or disordered motion of matter (the gas) within a rapidly moving medium (space itself).

Once the Matter Era dawned, all the atoms within the vast primordial clouds were set into motion, not only from the expulsion of the bang, but also from the heat of the fireball. The gas then had some “directed” kinetic energy—outward, from the ordered expansion of the Universe. It also had some “undirected” thermal energy—random, from the disordered aftermath of the blazing inferno. Intact pockets of gas undoubtedly surged this way and that, whirling here and shearing there amid collisions with one other, in addition to the disarrayed agitation of each of the individual atoms. Turbulence likely aided the growth of spinning eddies at those places where density fluctuations had already become established in the early Universe.

Turbulent eddies of this sort can be visualized by watching water swirl down the drain of a bathtub. In a sense, the swirling eddies themselves are turbulence. Even better examples can be created by moving your hand gently through water, or a teaspoon through coffee; swirling eddies naturally emerge in the wake of this turbulence. Water flowing past rocks in a stream also gives an appreciation for the whirlpools that naturally arise in its aftermath. Figure 2.11 might help to visualize such eddies.

FIGURE 2.11 — Eddies in water naturally occur behind a rock in a stream or a moving canoe paddle. Slowly moving your hand through the bathtub water produces similar phenomena.

Probably the best examples of the effects of turbulence are the fluffy clouds in Earth’s atmosphere. Especially vivid in photographs (see Figure 2.12) of the tops of clouds, taken with Earth-orbiting satellites, kilometer-sized eddies can be seen as density enhancements of atmospheric gas. Such whirling eddies are known to become more pronounced whenever air currents are particularly turbulent. Should they grow, in this case by accumulating additional amounts of moisture, the eddies may well form stormy depressions and occasionally even hurricanes hundreds of kilometers across.

FIGURE 2.12 — Satellite photographs of the top of Earth's cloud layers clearly show atmospheric eddies. Such swirling vortexes are especially vivid when strong surface winds encounter a mountain, creating air turbulence. This photograph was taken by a Gemini spacecraft above Tenerife, where the eddies result from the Atlantic winds hitting the Canary Islands. (NASA)

Here is a case, then, where studies of a terrestrial phenomenon—Earth’s weather—may help us understand one of today’s most vexing extraterrestrial problems. As suggested in Figure 2.13, planetary hurricanes roughly mimic the overall morphology, the pancake shape, the differential rotation, the disposition of energy within spiral galaxies, and even a hole or eye at the center. Though these two types of systems are entirely unrelated and of vastly different size, their many resemblances might teach us something about galaxy formation via the study of hurricane formation—a good example of an interdisciplinary inquiry. In particular, since most meteorologists agree that some sort of turbulent “priming” is needed to trigger hurricanes, the early stages of such storms could conceivably be used by astronomers to extract clues about the turbulence-enhanced density fluctuations that might have helped protogalaxies to emerge long ago.

FIGURE 2.13 — Terrestrial hurricanes share many similarities with spiral galaxies. At left is Hurricane Katrina in 2005, spanning about 1000 kilometers in the Gulf of Mexico. At right is M101, a grand spiral galaxy measuring ~1018 km (or ~100,000 light-years) across. (NOAO/AURA)

Gravitational Instabilities It’s worth pursuing this idea a little further, gearing the discussion again to the sequence shown in Figure 2.10. Despite the inevitable cooling caused by the expansion of the Universe, each localized eddy within a large gas cloud must begin to heat. It can’t avoid it. Eddies are sites, not only of turbulence, but also of rising heat within a steadily cooling cloud. The heat results from friction caused by frequent collisions among the increasingly dense collections of atoms within each eddy. The technical term is "collisional excitation," but the process is a simple one, not unlike the heat derived by rubbing hands together on a cold wintry day.

Where does the heat come from originally? The source is the cloud's gravitational potential energy. A common example is a pencil held high above a tabletop and said to have some gravitational potential energy relative to the tabletop; the pencil has even more potential energy relative to the more distant floor below. Potential energy—the energy of position—is a measure of the energy that could potentially be gained if an object fell a certain distance; if the fall is caused by gravity, we call it gravitational potential energy. While falling, some of that potential energy is changed into kinetic energy—the energy of motion. Upon hitting the tabletop, the pencil of course stops falling. There the pencil has no potential energy relative to the tabletop, since the two are not separated by any distance; nor does the pencil have any kinetic energy since it has no motion on the table. During this process, the original potential energy of the pencil transforms into kinetic energy while falling. That kinetic energy was in turn changed into thermal energy (heat) when the pencil collided with the tabletop. The heat created is not large, but a sensitive thermometer could be used to measure it. No energy disappears; it just changes from one type into another.

Similar reasoning can be applied to a gas cloud. The atoms near the edge of the cloud have some gravitational potential energy relative to the center of the cloud. Provided that the cloud is massive enough, the combined gravitational pull of each atom on all the other atoms will force the cloud to contract. As the atoms at the cloud's edge fall toward the cloud's center, their potential energy changes to thermal energy. In short, the infalling cloud heats up.

This newly gained heat causes the individual atoms to increase their motions. In this way, thermal energy tends to buoy a cloud, that is, to retard its contraction. To some extent, heat counteracts gravity. That's why the Sun remains a large ball of gas; the pressure of its great heat prevents gravity from collapsing it into a small ball.

The physical status of any gas cloud is mostly governed by the competition between its potential energy and its thermal energy. A cloud is unstable if its (inward-pulling) gravitational energy exceeds its (outward-pushing) thermal energy. Such a gravitational instability causes the cloud to contract. Some clouds can collapse rapidly if they’re cool enough. If the conditions are reversed—should the thermal energy of the cloud exceed its gravitational energy—the cloud will expand, becoming larger and thinner. Some clouds can disperse entirely if they’re hot enough. These opposing conditions can be expressed by two inequalities:

Contraction: gravitational potential energy > thermal energy

Expansion: gravitational potential energy < thermal energy.

All gas clouds tend to contract or expand until their gravitational energy equally balances their thermal energy. Such clouds have achieved dynamical equilibrium. Apart from minor fluctuations in the gas density at small places here and there, clouds in equilibrium will remain that way unless they are otherwise influenced—that is, until they are heated, cooled, or pushed around in some way.

Since our main goal here is to understand the origin of galaxies, let's suppose that the conditions for contraction are satisfied. A gravitational instability sets in, as the inward pull of gravity exceeds the outward buoyancy of heat. So the cloud infalls a little, its density increases, and some heating occurs. We might expect the newly generated heat to bring the cloud into equilibrium because, after all, the thermal energy would steadily increase until it equals the gravitational energy. If this were all that occurred, clouds everywhere would easily stabilize, neither galaxies nor stars could form—and we wouldn’t be here.

In reality, individual eddies eventually rid themselves of some of their newly acquired energy, much as the Sun or any other heated object needs to unload some energy, lest it blow up. The eddies in a protogalactic cloud do it by radiating away some of their heat. In this way, a large cloud containing lots of eddies can cool even faster than would a normally homogenous cloud in the expanding Universe. As it cools, the entire cloud contracts a bit more, thereby increasing the density and hence the heat within each eddy. Both the individual eddies and the whole cloud simultaneously radiate some of this newly gained energy into space, thereby allowing further contraction of the parent cloud and its smaller eddies. On and on, this cycle of contracting, heating, radiating, cooling, and contracting proceeds—all fundamentally driven by gravity. The cycle may operate at different speeds for each eddy, particularly since some eddies will be more successful than others at sweeping up additional gas from the parent cloud.

As sketched in Figure 2.14, it’s easy to conceptualize a cluster of galaxies forming in this way, with each eddy becoming a member galaxy within that cluster. Alternatively, perhaps only one or a few galaxies form within each of the vast primordial clouds of the early Matter Era, after which gravity gradually collects the galaxies into the very much larger galaxy clusters now seen scattered throughout the Universe. Either way, fragmentation models of this sort resemble a “top-down” approach to galaxy birth whereby gargantuan clouds give rise to litters of young galaxies—a process known in the trade as “monolithic collapse.”

FIGURE 2.14 — Eddies of gas within a large primordial cloud would have experienced a cyclical process of contracting, heating, radiating, cooling, contracting, and so on. Those eddies that survived the process presumably became galaxies.

A Complication If the vast primordial clouds had some rotation—and surely parts of them would have, at least at the sites of the swrling eddies—the eddies must have changed their shapes while contracting. Why? Because of the inertial tendency of matter in a rotating body to keep rotating. This is angular momentum, another one of those basic physical quantities, like matter and energy; specifically, it’s the tendency of an object to spin or move in a circle.

To appreciate how angular momentum can change the shape of a contracting object, consider first its simpler counterpart, linear momentum—the tendency of an object to keep moving in a straight line. A truck and a bicycle rolling equally fast down a street each has some linear momentum. Trying to stop them, you would obviously find it easier to halt the less massive bicycle. Although each vehicle has the same speed, the truck has more momentum. Thus the linear momentum of an object depends on the mass of that object. It also depends on the velocity, for if two bicycles rolled down the street at different speeds, you would be more likely able to stop the slower moving one. Consequently, linear momentum is defined as the product of both mass and velocity:

linear momentum = mass x velocity.

Angular momentum is only a little more complex, and it applies to objects having some rotation (on its axis) or revolution (about another object). In addition to mass and velocity, angular momentum depends on the size of the object:

angular momentum = mass x velocity x size.

(The velocity term for linear momentum is velocity in a straight line, whereas that for angular momentum refers to the circular velocity of spinning or orbital motion.)

Basic physics demands that both these momenta must be conserved at all times. In other words, momentum must remain constant before, during, and after a physical change within and among any objects. For example, if a spherical blob having some spin begins contracting, the relationship above demands that the circular velocity grow larger. After all, the blob's mass doesn’t change during contraction, yet its size clearly decreases. The circular velocity of the spinning blob must therefore increase in order to keep unchanged the total angular momentum—the product of mass, circular velocity, and size—at all times.

Figure skaters understand the principle of angular-momentum conservation. As shown in Figure 2.15, when the arms are drawn in, yet the mass of the human body remains the same, the lateral size decreases. Hence the circular velocity must increase in order to conserve angular momentum, and the result is that the skater spins more rapidly.

Figure 2.16 illustrates how continued contraction can force a cloud to rotate increasingly rapidly. And as the spin ramps up, the cloud gradually flattens—the greater the rotation, the flatter the cloud. That flattening occurs since matter contracts more easily along the axis of rotation than perpendicular to it. In all, the whole cloud becomes smaller, particularly parallel to its spin axis.

In some cases, angular momentum can mount enough to counteract gravity. Any rapidly spinning object has a tendency to force matter away from its center, much like mud thrown from a rotating bicycle wheel. So heat is not the only agent capable of opposing gravity. Rotation can do it too (as can magnetism sometimes). In fact, it’s rotation, not heat, that stabilizes galaxies once they’re fully formed.

The broad outline of this contraction-flattening sequence for galaxy formation is supported by observations. Young stars are spread throughout the plane of our Milky Way Galaxy; none of them resides in the halo. Only old stars are found in the halo of our Galaxy, as well as other galaxies. The monolithic collapse model suggests that galaxies were more spherical in their youth when their first stars were forming, after which the galaxies flattened somewhat into the shapes now seen.

FIGURE 2.15 — Figure skaters, like Katerina Witt shown here winning an Olympic Gold Medal, know well that the spin of their "pirouette" can be increased by tucking in their arms. That is, as her lateral size decreases yet her mass remains the same, her circular velocity must increase. And just as her skirt flew outward, the gas clouds that formed galaxies (or later our Solar System on much smaller scale) developed a disk perpendicular to its spin axis. (World Wide Photos)

FIGURE 2.16 — The conservation of angular momentum demands that a spinning and contracting gas cloud gradually flattens along its axis of rotation.

Another Complication As nice as this galaxy-formation scenario seems, it, too, runs into some serious problems once mathematics are applied to it. Calculations show that timing is once again an issue, but not, as above, that the eddies take too long to form. Here, it’s more a case of competing timescales between physical events affecting the eddies: The time needed for capture and contraction of the gas in a turbulent eddy is longer than the typical time for the random dissipation of that eddy. In other words, eddies tend to break up long before they have a real chance to bind tightly. Turbulent eddies do enhance the random gas fluctuations, but they don’t last long enough to form galaxies.

Any kind of eddy, then—in the bathtub or in the early Universe—comes and goes in iffy sorts of ways, all governed by the laws of statistical physics. Eddies appear, disappear, and reappear at different parts of either a terrestrial atmospheric cloud of moist air or an extraterrestrial galactic cloud of primordial gas. Occasionally, a terrestrial eddy does indeed grow to form a flourishing hurricane, or a primordial eddy presumably a genuine galaxy. But the expected rarity of their rapid growth implies that turbulent eddies cannot be the sole solution to the problem of the ancient formation of all of the galaxies or of galaxy-like objects.

Primeval Lumps Main-stream astrophysicists prefer to avoid radical theories of galaxy formation—such as a weird one postulating the ballooning of compact, primordial blobs (called by some, “white holes”) for which there’s no evidence whatsoever. They head back to first principles and embrace once again the basic notion of random gas fluctuations developing into something bigger—a “bottom-up” approach that groups smaller chunks of matter to build galaxies. Still, some additional means must be found to speed the growth of such fluctuations in the gas-radiation mix of the early cosmic fireball. Current research therefore centers on other ways that might have enhanced, or accelerated, the growth of simple gas fluctuations.

The general scenario now favored by the astronomical community—an idea known as hierarchical clustering—postulates an early Universe that was not homogeneous. Instead, it’s imagined to have been peppered, even in the PARTICLE EPOCH, with minute density clumps. In other words, the eddies got a head start even in the prior Radiation Era and thereafter acted as seeds for the growth of galaxies early in the Matter Era. These already-formed pockets of gas would then have further developed during the GALACTIC EPOCH to fabricate at least the essential bulk of galaxies seen today. Although this idea initially sounded like a cop-out to many astronomers, observational evidence for these truly primeval inhomogeneities was marvelously confirmed at the end of the 20^th century, allowing theorists to breathe a sigh of relief that they might be on the right track.

Our only direct probe into the early Universe is the cosmic background radiation noted near the end of the previous PARTICLE EPOCH. Launched at the end of the Radiation Era, ~300,000 years after the start of all things, its radio photons now engulfing us grant some inkling of the wild physical conditions prevailing at the time. Briefly explained, radiation is influenced by the gravity of growing clumps of matter, so that as the density of the clumps varied from place to place in the early Universe, the observed radiation—then launched and now observed—ought to show slight temperature variations from place to place on the sky. Such “ripples” in the temperature of the background radiation have indeed been spotted, though only weakly at the level of parts per million. That is, given that the average temperature of the fossil radiation is ~3 K, the minute thermal variations (Figure 2.17) that have been detected by radio receivers aboard Earth-orbiting satellites, most notably the Wilkinson Microwave Anisotropy Probe (WMAP), are only on the order of millionths of one degree. Yet, they are in accord with those expected for a wide range of theoretical models of galaxy formation, including the superclusters, voids, filaments, and bubbles observed all across the firmament.

FIGURE 2.17 — This map of temperature fluctuations in the cosmic background radiation was produced by the WMAP satellite in Earth orbit. It’s much more sensitive than the one shown earlier in the PARTICLE EPOCH (Fig 1.13 ), here depicting across the entire sky hotter than average regions in red and cooler than average regions in blue. The temperature changes are minute, spanning only millionths of a kelvin. Those differences imply regions of enhanced density and are thought to represent the “seeds” from which galaxies began forming once the Matter Era began, ~500,000 years after the big bang. (NASA)

Hierarchical Clustering Here, in a nutshell, is the basic idea of hierarchical clustering, considered more of an ongoing process than a single event: Extremely small-scale fluctuations in the matter density present before the time of inflation—an inevitable consequence of quantum physics operating in the very early Universe well less than a second old—would have been stretched and amplified by inflation to a size and scale typifying whole galaxies and even larger. The subsequent growth of those gravitational instabilities, already established when the Radiation Era gave way to the Matter Era, probably led to the gradual formation of self-gravitating collections of matter. Should this idea be correct, then the vast assemblages of matter seen today as galaxies, galaxy clusters, and even the gigantic galaxy superclusters are the progeny of subatomic quantum effects prevalent when the Universe was a mere 10^-35 second old.

The accepted mechanism of galaxy formation, still only roughly understood, is then a familiar one to experts of star formation, as will be examined in detail in the next STELLAR EPOCH. Nature quite naturally selects mass-density fluctuations that gravitationally induce cycles of contracting, heating, radiating, cooling, and eventual flattening into disk-shaped objects. But, just when we feel good about getting closer to grasping reality, another complication sets in. For galaxies yet unlike for stars, these events didn’t likely involve only normal matter; dark matter has been implicated to some (unknown) extent and that clearly confuses things.

Given the prevailing conditions in the early Universe, specifically at the interface of the Matter and Radiation Eras, only regions of higher-than-average density containing more than ~10⁶ times the mass of the Sun would have begun to contract. However, if galaxies grew long ago exclusively from the fluctuations within normal matter (that is, in the absence of dark matter), those density fluctuations should manifest themselves now as a clear observable imprint of large temperature variations in today’s cosmic background radiation; that imprint is not observed.

Instead, if dark matter was involved, it might have acted as that long-sought agent, or gravitational scaffolding, to help normal matter clump earlier in the Universe (when we know the bulk of most galaxies did form). The reason is that dark matter—whatever its true nature—interacts only weakly with normal matter and with radiation. So, its natural tendency to gravitationally infall (for dark matter still exerts gravity) was neither hindered by radiation, nor expected to leave a large signature on the cosmic background radiation. Accordingly, dark matter, being ~5 times more abundant than normal matter, probably clumped first and then acted as an accelerant to draw normal matter into the regions of highest density. This scenario explains why so much dark matter seems to reside in the vicinity of the visible galaxies: That’s where the dark matter initially concentrated, thus attracting the normal matter that became the galaxies now so luminously seen. The brightly lit galaxies resemble the visible tips of mostly hidden icebergs, or the illuminated bulbs on an otherwise dark christmas tree.

Of course, all this modeling is a little shaky given that astronomers don’t yet know the nature of that dark matter. To be honest, some of the uncertainty is welcome, allowing theorists much freedom in choosing the dark matter’s properties while seeking to match galaxy-formation models with observed structure on the sky—and that, in turn, might imply valuable information about the dark matter. To be just as honest, the theorists may be—to make a bad pun—whistling in the dark, as their models depend on vast quantities of abnormal matter that is only inferred and has never been detected.

In any case, the seeds of galaxy formation were likely sown in the early Universe when small density fluctuations in the primordial matter began to grow. The initial masses of these pregalactic blobs were quite small by galactic standards—perhaps only a few million, yet more likely a few billion, solar masses, comparable to those of the smallest, irregular galaxies. Those dwarf galaxies now seen scattered throughout galaxy clusters may well be the building blocks of galaxies—the so-called “baby galaxies.” As we shall see in the final section of this GALACTIC EPOCH , a growing consensus champions the idea that today’s big galaxies formed by the repeated merging and accumulating of smaller objects. This is indeed a “bottom-up” scheme, but not one that begins with objects as small as stars and planets; rather with million-to-billion solar-mass blobs that emerged near the start of the Matter Era.

Bilateral Support Support for this hierarchical scenario is moderate and on two fronts. Theoretical backing is provided mainly by computer simulations stipulating how normal (baryonic) and abnormal (dark) matter might have interacted with radiation during the first few billion years. These computer models demonstrate that merging of smaller into bigger objects was a viable phenomenon in the GALACTIC EPOCH and could have conceivably led to the formation (and evolution) of the many varied galaxies observed today. Although the models have wide latitude among their input parameters, while at the same time suffer from computer codes obviously not as robust as the real cosmos, no “show-stoppers” have yet intruded—nothing in the theoretical analyses that leads us to believe we are not on the right track, finally.

Observational support for hierarchical clustering has also increased lately, especially recent findings that some of the most remote galaxies (namely, those seen in their youth) appear somewhat smaller and less regular than those found nearby. Deep, long-exposure images acquired by the world’s most powerful telescopes—such as the Hubble Space Telescope in orbit, the Keck Observatory in Hawaii, and the Very Large Telescope in Chile—show evidence for distant, distorted spheroids containing a million to a billion solar masses (but no distinct stars) in regions typically a few thousand light-years across—roughly the size and scale expected for pregalactic building blocks. We seem to be seeing these blobs as they were ~12 billion years ago, perhaps poised to merge into larger, galaxy-sized objects. Alas, not all astronomers buy this interpretation, as the data are sketchy, the images fuzzy, and the modeling simplified. It remains unclear if anyone has yet seen a genuine “baby” galaxy or any luminous object caught in the act of galactic birth—another of science’s unachieved grails.