Chapter 9

Our Solar System and Others

Introduction Astronomers have long speculated about the

origin of our Solar System. They have noted regularities in the way the

planets orbit the Sun and in the spacing of the planetary orbits.

But until recently, astronomers have been limited to studying one planetary system: our own.

In 1600, Giordano Bruno was burned, naked, at the stake for wondering about the plurality of worlds, to use the term of that day.

Finally, after hundreds of years of wondering whether they existed, another planetary system was discovered in 1991.

And in the last decade, planetary systems galore were discovered around stars like the Sun.

Introduction While in 1990 we knew of only 9

planets, all around our Sun, as of this writing (late-2005) we know of more than 160 additional planets around other stars (see figure).

In this chapter, we will first discuss our own Solar System and its formation.

Then we will describe how we find other planetary systems and how they may have been formed.

This discussion, in turn, should give us insights into how well we understand the origin of our own Solar System.

9.1 The Formation of the Solar System

Many scientists studying the Earth and the planets are particularly interested in an ultimate question:

How did the Earth and the rest of the Solar System form?

We can accurately date the formation by studying the oldest objects we can find in the Solar System and allowing a little more time.

For example, astronauts found rocks on the Moon older than 4.4 billion years.

We have concluded that the Solar System formed about 4.6 billion years ago.

9.1a Collapse of a Cloud

Our best current idea is that 4.6 billion years ago, a huge cloud of gas and dust in space collapsed, pulled together by the force of gravity.

What triggered the collapse isn’t known.

It might have been gravity pulling together a random cloud, or it might have been a shock wave from a nearby supernova (see figure).

As the gas pressure increased, eventually the rate of collapse decreased to a slower contraction.

9.1a Collapse of a Cloud

You have undoubtedly noticed that ice skaters spin faster when they pull their arms in (see figure).

Similarly, the cloud that ultimately formed the Solar System began to spin faster as it collapsed and contracted.

The original gas and dust may have had some spin, and this spin is magnified in speed by the collapse, because the total amount of “angular momentum” doesn’t change.

It spins faster because the quantity known as angular momentum stays steady in a rotating system, unless the system is acted on by an outside force (assuming the force is not directed at the center of the object).

The amount of angular momentum depends on the speed of the spin and on how close each bit of spinning mass is to the axis of spin; see our discussion at the end of Chapter 5.

9.1a Collapse of a Cloud Objects that are spinning around tend to

fly off, and in our case this force eventually became strong enough to counteract the effect of gravity pulling inward.

Thus the Solar System stopped contracting in one plane.

Perpendicular to this plane, there was no spin to stop the contraction, so the solar nebula ended up as a disk.

The central region became hot and dense, eventually becoming hot enough for the gas to undergo nuclear reactions (see Chapter 12), thus forming our Sun.

9.1a Collapse of a Cloud

In the disk of gas and dust, we calculate that the dust began to clump (see figures).

Smaller clumps joined together to make larger ones, and eventually planetesimals, bodies that range from about 1 kilometer up to a few hundred kilometers across, were formed.

Gravity subsequently pulled many planetesimals together to make protoplanets (pre-planets) orbiting a protosun (pre-Sun).

9.1a Collapse of a Cloud The protoplanets then contracted and cooled to

make the planets we have today, and the protosun contracted to form the Sun (see figure).

Some of the planetesimals may still be orbiting the Sun; that is why we are so interested in studying small bodies of the Solar System like comets, meteoroids, and asteroids.

Most of the unused gas and dust, however, was blown away by a strong solar wind.

9.1b Models of Planet Formation

In one of the main models for the formation of the outer planets, the solar nebula first collapsed into several large blobs.

These blobs then became the outer planets (see figure, right).

The two outermost blobs lost most of their gaseous atmospheres as a result of strong ultraviolet light from the young Sun that also removed the gas from the nebula at Saturn’s orbit and beyond (see figures, below).

As a result, they became the moderately massive planets, Uranus and Neptune.

Because of the strength of their internal gravity, Saturn lost only a portion of its gas, while Jupiter was unaffected, thus becoming the most massive planet.

9.1b Models of Planet Formation In another model, a solid core (resembling a

terrestrial planet) condensed first for each of the outer planets.

The gravity of this core then attracted the gas from its surroundings.

For this second model, the relative amounts of elements in the rocks and in the gases would also differ from planet to planet.

Spacecraft found that Jupiter, Saturn, Uranus, and Neptune have different relative amounts of some of the elements in their atmospheres.

Additionally, the atmospheres of Jupiter and Saturn are very much more massive than the atmospheres of Uranus and Neptune.

In this model, the cores of Uranus and Neptune may contain 10 to 15 times the mass of the Earth.

Also in this model, Jupiter’s core is thought to have only 0 to 3 Earth masses, much less than the cores of Uranus and Neptune, which is an argument against this model.

9.1b Models of Planet Formation In the inner Solar System, the terrestrial

planets are the accumulation of planetesimals. Our Earth and its neighbor planets were formed

out of planetesimals made of rocky material. The rocky material, mainly silicates, condensed

at the temperatures of these planets’ distances from the Sun and were not blown outward by particles from the forming Sun.

These terrestrial planets never became massive enough to accumulate a massive atmosphere the way the giant planets did.

And, because the inner planets are closer to the Sun and therefore hotter, gas in their atmospheres moved relatively fast.

Thus free hydrogen and helium escaped from the Earth’s low gravity, while Jupiter and the other giant planets have huge atmospheres of hydrogen and helium, matching the atmosphere of the Sun.

9.1b Models of Planet Formation Early on, the Sun would have been spinning very

fast. But much of the excess angular momentum was transported upward by a “bipolar outflow”; see the discussion in Chapter 12.

In 2001, John Chambers and George Wetherill of the Carnegie Institution of Washington proposed a model extending standard ideas of Solar-System formation to explain not only the range of planets we see but also the relative emptiness of the asteroid belt.

Many people mistakenly think that the asteroids are the remains of a large planet that exploded, but the asteroid belt actually contains so little material that not even a small moon could have been present there.

In their model, planetesimals formed everywhere throughout the solar nebula, including the asteroid belt.

Some of the planetesimals coalesced into planets.

9.1b Models of Planet Formation Jupiter grew especially rapidly, and its gravity

kicked out material in the asteroid-belt region that orbited an integer number of times in the time that Jupiter itself made one orbit (or some other integer number of orbits).

This aspect of Jupiter’s gravity, and that it formed gaps in the asteroid belt, has long been known.

Chambers and Wetherill have added the idea that objects in this region that are not quite in the places affected so strongly by Jupiter’s gravity were pushed into these zones by gravitational encounters among themselves, including some planetesimals and small planets that had formed there.

They have checked the idea with computer simulations. The process also made the orbits of the giant

planets more circular, matching observations. Further, it sent some objects that would contain

volatile substances (those that evaporate easily) like water to crash into the Earth.

This process may explain how we got our oceans, something more often attributed to comets.

9.1b Models of Planet Formation It is interesting that nothing about our

current model of planetary-system formation implies that the Solar System is unique.

As we will see next, we are increasingly finding systems of planets around other stars.

To our surprise, their properties aren’t like those of our Solar System: We can’t see if they have small, rocky planets close in to the star, but we know that they don’t all have massive ones farther out.

Maybe some of our modelling for our own Solar System has been wrong because we looked for regularity in the distribution of planets while we really had a random distribution.

9.2 Extra-solar Planets (Exoplanets)

People have been looking for planets around other stars for decades.

Many times in the last century, astronomers reported the discovery of a planet orbiting another star, but for a long time each of these reports had proven false.

Finally, in the 1990s, the discovery of extra-solar planets (planets outside, “extra-,” the Solar System) seemed valid.

They have also become known as exoplanets.

9.2a Discovering Exoplanets Since exoplanets shine only by reflecting a

small amount of light from their parent stars (i.e., the stars that they orbit), they are very faint and extremely difficult to see in the glare of their parent stars with current technology.

So the search for planets has not concentrated on visible sightings of these planets.

Rather, it has depended on watching for motions in the star that would have to be caused by something orbiting it.

9.2a(i) Astrometric Method

The earlier reports, now rejected, were based on tracking the motion of the nearest stars across the sky with respect to other stars.

The precise measurement of stellar positions and motions is called astrometry, so this method is known as the astrometric method.

If a star wobbled from side to side, it would reveal that a planet was wobbling invisibly the other way, so that the star/planet system was moving together in a straight line. (Technically, the “center of mass” of the system has to move in a straight line, unless its motion is distorted by some outside force. The center of mass is illustrated in the figure, and is described more fully in Chapter 11, when we discuss binary stars. Both the star and the planet orbit their common center of mass, though the star is much closer to the center of mass than the planet is. Thus, the star moves very slightly, in a kind of “reflex motion” caused by the orbiting planet.)

9.2a(i) Astrometric Method Astrometric measurements have been

made over the last hundred years or so, and a few of the nearby stars whose motions in the sky were followed seemed to show such wobbling.

But the effects always turned out to be artifacts of the measuring process.

Nevertheless, the astrometric method is still being used by some astronomers, and maybe they will eventually detect an exoplanet with it.

9.2a(ii) Timing of Radio Pulsars The first extra-solar planet was discovered

in 1991 around a pulsar, a weird kind of collapsed star (see our discussion in Chapter 13) that gives off extremely regular pulses of radio waves with a period that is a fraction of a second.

The pulses came more frequently for a time and then less frequently in a regular pattern.

So the planet around this pulsar must be first causing the pulsar to move in our direction, making the pulses come more rapidly, and then causing it to move in the opposite direction, making the pulses come less often. (As described above, the planet and star are actually orbiting their common center of mass.)

9.2a(ii) Timing of Radio Pulsars But this planet must have formed after the

catastrophic explosion that destroyed most of the star after its inner core collapsed; the planet could not have survived the stellar explosion.

Thus, it was not the kind of planet that is born at the same time as the star, like Earth.

Even when the existence of two more planets (and possibly a fourth planet) orbiting that pulsar was established, all with masses comparable to those of the terrestrial planets in our Solar System, the pulsar system seemed too unusual to think much about, except by specialists.

It didn’t help that another pulsar planet, discovered slightly earlier, turned out to be a mistaken report.

9.2a(iii) Periodic Doppler Shifts

In the 1990s, techniques were developed using the Doppler effect.

Recall that Chapter 2 describes how the Doppler effect is a shift in the wavelengths of light that has been emitted, caused by motion of the source or the receiver along the line of sight (that is, by the “radial velocity”).

The planet has a large orbit around the center of mass, moving rapidly.

But the parent star, being much more massive than the planet, is much closer to the center of mass (see figure, right) and therefore moves much more slowly in a smaller orbit.

With sufficiently good spectrographs and numerous observations, this slight “wobble” can be detected as a periodically changing Doppler shift in the star’s spectrum (see figure, left); the radial velocity of the star varies in a periodic way.

9.2a(iii) Periodic Doppler Shifts The breakthrough came because new computer

methods were found that measure the changing Doppler shifts very precisely.

On a computer, the star’s spectrum can be simulated including changes in the wavelengths of light, just as happens in the Doppler effect.

These simulated spectra can be compared to the observed spectrum of that star, until an excellent match is found.

This method allows very small Doppler shifts to be detected, and the speeds of stars toward or away from us can be measured to a precision of 1 meter/sec, a leisurely walking speed. (This very high precision was only recently achieved, with the Keck-I telescope; at most other sites, the precision has typically been 3 meters/sec or worse.)

9.2a(iii) Periodic Doppler Shifts The first surprising report came in 1995

from a Swiss astronomer and his student, Michel Mayor and Didier Queloz.

They found a planet around a nearby star, 51 Pegasi.

One strange thing about the planet is that it seemed to be a giant planet, at least half as massive as Jupiter, but with an orbit far inside what would be Mercury’s orbit in our Solar System.

The planet orbited 51 Peg in only 4.2 days, much faster than any planet in our Solar System.

9.2a(iii) Periodic Doppler Shifts Two American astronomers, Geoff Marcy and Paul

Butler, then at San Francisco State University and the University of California, Berkeley, had been collecting similar data on dozens of other stars.

But they had assumed, reasonably, that a planet like Jupiter around another star would take years to orbit, so they were collecting years of data while perfecting their analysis techniques to measure exceedingly small speeds.

They hadn’t run their spectra through the computer programs they were writing to measure the Doppler shifts.

When they heard of the Mayor and Queloz results, they quickly examined their existing data and also observed 51 Peg.

They soon verified the planet around 51 Peg and discovered planets around several other stars (see the drawing opening this chapter).

9.2a(iii) Periodic Doppler Shifts

Most of the new objects turned out to be giant planets either in extremely elliptical orbits or in circular orbits very close to the parent stars (see figure).

Most of these planets are orbiting stars within 50 light-years from us, not extremely far away but not the very closest few dozen stars either.

Stars this close are bright enough for us to carry out the extremely sensitive spectroscopic measurements.

9.2a(iii) Periodic Doppler Shifts

One limitation of the method is that we generally don’t know the angle of the plane in which the planets are orbiting their parent stars.

The Doppler-shift method works only for the part of the star’s motion that is toward or away from us, and not for the part that is from side to side.

So the planets we discover can be more massive than our measurements suggest; we are able to find only a minimum value to their masses (see figure).

9.2a(iii) Periodic Doppler Shifts In the first few years of exoplanet discovery,

this problem left the nagging question of whether the objects were really planets or merely low-mass companion stars.

They might even be objects called “brown dwarfs,” which have between about 10 and 75 times Jupiter’s mass, not quite enough to make it to “star status” (see our discussion in Chapter 12, and in Section 9.2c below).

Nevertheless, most astronomers believed these objects are planets, because there is a large gap in mass between them and lowmass stars.

Very few intermediate-mass objects (10 –75 Jupiter masses) had been found, yet they should have been easily detected if they existed.

This gap suggested that the new objects are much more numerous than brown dwarfs.

9.2a(iii) Periodic Doppler Shifts The discovery, in 1999, of a system of three

planets around the star Upsilon Andromedae clinched the case that at least most of the objects are planets.

Other multiple-exoplanet systems were found thereafter. It seems most unlikely that a system would have

formed with four closely spaced stars (or brown dwarfs) in it, while a system with one star and three planets is reasonable.

One of Upsilon Andromedae’s planets even has an orbit that corresponds to Venus’s in our Solar System, not as elliptical as the orbits of the planets around other stars.

This planet is in a zone that may be not too hot for life nor too cold for it.

Though such a massive planet would be gaseous, and so not have a surface for life to live on, it could have a moon with a solid surface, just as the giant planets in our Solar System have such moons.

9.2a(iii) Periodic Doppler Shifts

However, a considerable advance was announced in 2005: the existence of a planet whose mass is as low as only 7 to 8 times Earth’s (see figure), by a group of scientists including Geoff Marcy and Paul Butler.

This exoplanet orbits a star, known as Gliese 876 or GJ876, that is only 15 light-years from Earth.

Astonishingly, the planet orbits its parent star with a period of only two days, meaning that it is only 1/50 of an A.U. out, just 10 times the star’s radius.

It is so close to its star’s surface that its temperature is probably 200°C to 400°C, so it isn’t a candidate for bearing life as we know it.

Being so hot, yet having relatively low mass and thus moderately weak gravity, it could not have retained a lot of gas.

It is therefore apparently the first rocky, terrestrial-type planet ever found orbiting another star.

The exciting announcements continue. We now know of several systems that each

contain a few planets. Our methods are still not sufficiently

sensitive to find “minor” bodies like our own Earth.

9.2a(iii) Periodic Doppler Shifts

Finding this planet required improving the Keck telescope’s system for detecting small Doppler shifts (see figure, left).

The detailed observations (see figure, right) allowed a computer model to account for the angle at which we are viewing the system, which turns out to be inclined to our line of sight by 40°.

That measurement allowed the mass of the orbiting planets themselves to be determined, and not merely lower limits as before.

The measurements revealed that the slight discrepancy in the orbits of the two already-measured planets could be resolved by the presence of the third body, the newer exoplanet.

9.2a(iv) Transiting Planets

Since 1999, astronomers have not had to resort only to periodically changing Doppler shifts to detect exoplanets.

One planet was discovered to have its orbit aligned so that the planet went in front of the star each time around—that is, it underwent a transit.

The dip in the star’s brightness of a few per cent can be measured not only by professional but also by amateur astronomers (see figures).

9.2a(iv) Transiting Planets Since the planet’s orbital plane is along our

line of sight, its mass can be accurately determined, and this turns out to be 63 per cent of Jupiter’s mass.

This result confirms our conclusion that at least some (and probably most) of the “exoplanets” really are planets rather than more massive brown dwarfs.

The transit method has also revealed, through spectroscopy, sodium in the exoplanet’s atmosphere.

The atmosphere produced some absorption lines in the starlight passing through it, and these were detected in very high-quality spectra.

In the future, it is possible that astronomers will detect evidence for life on other planets by analyzing the composition of their atmosphere using the transit method.

9.2a(iv) Transiting Planets During the past few years, several additional

examples of transiting planets have been found. Many more exoplanets will be discovered in this

way from the ground and from space. This method is analogous to observing the

transit of Venus across the disk of our Sun (see figure).

Progress in the search for exoplanets is so rapid that you should keep up by looking at the Extrasolar Planets Encyclopaedia at http://www.obspm.fr/planets and the Marcy site at http://exoplanets.org, both linked through this book’s web pages.

9.2b The Nature of Exoplanet Systems

We have discovered enough exoplanets to be able to study their statistics.

About 1 per cent of nearby solar-type stars have jovian planets in circular orbits that take between 3 and 5 Earth days.

These are sometimes called “hot Jupiters,” since they are so close to their parent stars (within 1/10 A.U.) that temperatures are very high.

Another 7 per cent of these nearby stars have jovian planets whose orbits are very eccentric.

As we observe for longer and longer periods of time, we have better chances of discovering planets with lower masses or with larger orbits.

9.2b The Nature of Exoplanet Systems

The discovery of several planetary systems instead of just our own will obviously change the models for how planetary systems are formed.

Since giant planets couldn’t form in the torrid conditions close in to the parent star, theorists work from the idea that the exoplanets formed far out, as Jupiter did, and migrated inward.

Thus it follows that these planets may be jostled loose from their orbits and put in orbits that bring them closer to their parent stars.

Maybe their orbits shrink as the planets encounter debris in the dusty disk from which they formed, or shrink along with an overall swirling inward of the whole disk of orbiting material.

9.2b The Nature of Exoplanet Systems

Perhaps the highly elliptical orbits didn’t start out that way, but were produced by gravitational interactions that completely ejected some planets from the system.

The gravitational interactions between two planets can lead to the ejection of one planet, leaving the other in a very eccentric orbit.

Another idea is that interactions between a planet and the protoplanetary disk can cause high eccentricities.

Once a planet has part of its orbit close to its parent star, tidal forces can circularize the orbit.

Other planets can spiral all the way into the star and be destroyed.

9.2b The Nature of Exoplanet Systems

The most accepted explanation for the hot Jupiters, which orbit so close to their stars, is that they were formed farther out and migrated in.

But the 2005 discovery of a hot Jupiter in a triple star system complicates matters, since the two farther-out stars would have disrupted any protoplanetary disk.

So this discovery is being interpreted as evidence against the migration model.

Perhaps that model was partly based on residual prejudice that our own Solar System’s outer giant planets are normal.

Perhaps the triple system had a very different type of protoplanetary disk than the one with which we are familiar from our own system, or the planet was captured.

9.2c Brown Dwarfs As mentioned in Section 9.2a(iii), some of the objects

found with the Doppler shift technique might actually be too massive to be true “planets” (more than 13 Jupiter masses).

If so, they are brown dwarfs, which are in some ways “failed stars.”

Each of the brown dwarfs has less than 75 Jupiter masses (or 7.5 per cent the mass of the Sun), which is not enough for them to become normal stars, shining through sustained nuclear fusion of ordinary hydrogen, as we shall discuss in Chapter 12.

Their central temperatures and pressures are just not high enough for that. (However, they do fuse a heavy form of hydrogen known as “deuterium,” so they are not complete failures as stars.

We discuss the origin of deuterium, all of which was formed in the first few minutes after the Big Bang, in Chapter 19.)

Brown dwarfs can be thought of as the previously “missing links” between normal stars and planets.

9.2c Brown Dwarfs Though no detection of a brown dwarf was accepted

until 1995, hundreds have now been observed. Many of these are in the Orion Nebula, while

others are alone in space. The current best model for brown dwarfs is that

they are formed similarly to the way that normal stars are: in contracting clouds of gas and dust.

A disk forms, perhaps even with planets in it, though the material in it contracts onto the not-quite-star.

This idea is backed up by observations with the European Southern Observatory’s Very Large Telescope, which has detected an excess of near-infrared radiation from many brown dwarfs.

The scientists involved interpret their observations as showing that the radiation is from dusty disks.

Further, they conclude that since both regular stars and brown dwarfs have such disks, they must also form in similar ways.

9.2c Brown Dwarfs Most exciting, a planet next to a brown dwarf

(presumably orbiting it) has been imaged (see figure), the first planet to be imaged around any star.

The star with its planet is 200 light-years away from us; observations from the ground and from space have shown that the two objects are moving through space together.

The planet is 55 A.U. away from its parent star.

9.2d Future Discovery Methods

NASA’s Kepler mission, planned for a 2008 launch, is to carry a 1-m telescope to detect transits of planets across stars.

To do so, it will continuously monitor the brightness of 100,000 stars. The 2004 transit of Venus

served as an analogue to the type of thing Kepler will study: One of this book’s authors (J.M.P.) and colleagues reported on the dip in the total amount of sunlight reaching Earth because Venus was blocking about 0.1 per cent of the Sun’s disk (see figure).

9.2d Future Discovery Methods No telescope now in space has enough

resolution to directly detect (image) a planet closely orbiting a normal star (rather than a brown dwarf as described in the preceding section).

It will take an interferometry system, with two widely separated telescopes working together, to make such a detection.

Installation of interferometric equipment at the two 10-meter Keck telescopes in Hawaii and at the four 8-meter units of the Very Large Telescope in Chile is nearing completion, which should lead to direct detection of some young, giant planets that are luminous at infrared wavelengths.

9.2d Future Discovery Methods

Though we have mainly detected giant planets, we are now beginning to detect small planets in those systems.

To do even better, NASA’s Space Interferometry Mission, now called SIM PlanetQuest (see figure), is on the drawing board, but it will not be launched until at least 2011.

NASA’s Terrestrial Planet Finder is to be able to image small planets and is slated for launch in the third decade of this century.

In 2005, both were delayed by an unfortunate shift in NASA’s priorities.

9.2d Future Discovery Methods

For these missions, NASA is currently examining two approaches to high-contrast imaging.

The first would use a huge telescope for direct imaging, though blocking the starlight itself to reveal accompanying planets.

The second would use several infrared telescopes flying in formation and coupled to form a “nulling interferometer,” in which the response at the stellar position is minimized (see figure).

NASA is considering an earlier, smaller, less expensive mission in its Discovery class of spacecraft to try out the technologies.

9.2d Future Discovery Methods The European Space Agency (ESA) also plans

spacecraft to find exoplanets. But again, funding reasons have delayed or abandoned their plans.

In 2004, they cancelled their Eddington mission to search for Earth-like systems by looking for their transits.

Their Gaia mission is to measure positions for a billion stars, and it may discover 10,000 planets!

It should be launched by 2012. ESA’s Darwin mission is to analyze the

atmospheres of Earth-like planets to search for signs of life, but not until at least 2015.

It is to consist of a flotilla of three 3-m telescopes on spacecraft in formation.

9.2d Future Discovery Methods Also, stars are being monitored to see if

their gravity focuses and brightens the light from other stars behind them, a process called “gravitational lensing” that we will discuss in Chapters 16 and 17.

The hope is that not only a background star but also, slightly before or after the star passes, a planet will cause a brightness blip.

Some candidate events have been reported, but more work is necessary before they will be widely accepted as true evidence of exoplanets.

9.3 Planetary Systems in Formation

We are increasingly finding signs that planetary systems are forming around other stars.

One of the first signs was the discovery of an apparent disk of material around a southern star in the constellation Pictor (see figures, right).

The best observations of it were made with the Hubble Space Telescope, and may show signs of orbiting planets.

An even nearer planetary disk has been found, enabling observations with higher resolution (see figure, left).

9.3 Planetary Systems in Formation

Other images of regions in space known as “stellar nurseries” show objects that appear to be protoplanetary disks (see figures, top).

Observations reported in 2005 show that these objects contain about as much mass as a planetary system, clinching the idea that they are locations where planets are forming.

Locations where there are planets in formation glow in the infrared (see figure, bottom), because of the wavelength of the peak of the black-body curves for the temperature of warm dust (see our discussion of black bodies in Chapter 2); thus, the Spitzer Space Telescope and, much later, the Webb Space Telescope, should give many insights into planetary formation.

9.3 Planetary Systems in Formation The Hubble Space Telescope has imaged such a

ring of dust around the nearby star Fomalhaut (see figure), recording signs that a planet is tugging on it gravitationally.

9.3 Planetary Systems in Formation A Wesleyan University team has found a young

Sun-like star that winks on and off, apparently being eclipsed by dust grains, rocks, or asteroids that are orbiting it in a clumpy disk.

The star is near the Cone Nebula, a prolific nursery of young stars.

It fades drastically over 2.4 days to only 4 per cent of its maximum brightness, stays dim for another 18 days, and then brightens for another 2.4 days out of every 48.3 days.

No single object could provide such a long eclipse, so only an orbiting collection of smaller objects seems to match the observations.

The actual eclipse could be from a wave of gas and dust triggered by the masses of these objects.

The system, only 3 million years old, is changing from month to month.