NGC 2207 and IC 2163


Below left: Seemingly dangling from the Big
Dipper’s Handle, the Whirlpool Galaxy (M51, or
NGC 5194) in Canes Venatici is the spiral gal-
axy’s spiral galaxy. Yet it remains enigmatic,
with some astronomers attributing its signa-
ture spiral to a tidal pull from its companion,
NGC 5195 (at the tip of M51’s northern arm),
and others asserting that the spiral predated
any interaction between the paired galaxies.

Left: The Whirlpool’s dust lanes glow promi-
nently in this image from the now-defunct
Infrared Space Observatory. Maps like these
may help astronomers discern the interplay
between spiral arms and star formation.


©2002 Sky Publishing Corp. All rights reserved.
The beautiful patterns in spiral galaxies
continue to defy easy explanation.
By Jack Lucentini

A lar Toomre is annoyed. Despite a long, successful career in which he has
contributed much to today’s understanding of spiral galaxies, the MIT
professor admits a failure needles him. He hasn’t been able to make a
computer simulation showing how a famously beautiful galaxy, M51, de-
veloped its unusually symmetrical spiral shape, which he suspects is due
to a companion galaxy’s tidal pulls.

“The broad answer to why galaxies “It must be one of the oldest unsolved
form spiral patterns has become almost problems in astronomy,” says dynamicist
certainly clear in recent decades. Gravity Jerry Sellwood (Rutgers University). Other
does it,” says Toomre. “But there are researchers, like Toomre, contend that
some annoying details left. M51 has been the problem is mostly solved. The trou-
an irritant . . . it’s embarrassing to me.” ble is, they offer differing solutions.
Above: In 1845 William Parsons, the Third Earl
The pretty spiral patterns that charac- Spiral-galaxy aficionados are gradually
of Rosse, was the first to record M51’s spiral
terize some two-thirds of known galaxies finding consensus on some points,
structure, which he discerned at the eyepiece
— including ours — have produced em- though, and they also have produced
of his 72-inch reflector, then the world’s
barrassments not only for Toomre but some intriguing new theories in the past
largest. From The Scientific Papers of William
also for modern astronomy. Even as as- few years. These range from proposals
Parsons, Third Earl of Rosse.
tronomers throughout the past century that spirals constantly reshape them-
Facing page, right: Hubble Space Telescope have lustily attacked mind-boggling selves, like flags flapping in the breeze, to
images suggest that acoustic pressure waves questions such as how the universe spent a theory that galaxies organize them-
— not the canonical gravity-based density its first second, they have struggled with selves into spirals because it helps them
waves — operate within the Whirlpool’s in- the seemingly basic question of how fulfill their ultimate destiny: to collapse
nermost 1,000 light-years (circled region). these spirals originate and behave. into disorderly lumps.

©2002 Sky Publishing Corp. All rights reserved. Sky & Telescope September 2002 37
Not until the 1920s were galaxies even recognized for what they are —

First Steps astronomers’ satisfaction. ent stars at different times. Lindblad
In 1845 Ireland’s Earl of Rosse first rec- Another early and obvious concept never could explain precisely how these
ognized spiral structure in Messier 51 was that the arms are material entities — patterns would operate or originate, but
(M51), then called the Whirlpool Nebu- long trails of stars, basically. As the gal- the idea stimulated future thinking.
la. Such objects, he wrote, “awaken an in- axy rotates (and spiral galaxies long have
tense desire to learn something of the been known to do so), the arms and stars Density Waves Debut
laws which give order to these wonderful follow in step, according to this view. But In the 1960s MIT’s C. C. Lin and Frank
systems.” Swedish astronomer Bertil Lindblad saw Shu fleshed out the hypothesis into an
Yet it wasn’t until the 1920s that galax- a problem with this. Since galaxies rotate influential theory that spiral arms are
ies even were recognized for what they differentially — the material near the “density waves.” Density waves are travel-
are — giant agglomerations of stars, center completes an orbit faster than the ing zones of compression, areas where
similar to and outside of our Milky Way outermost stars and gas clouds — mater- matter is more tightly packed than in the
galaxy. By the 1950s, astronomers were ial spiral arms would wind up tightly surroundings. The compaction moves
seriously venturing into the spiral-arm around the hub, becoming unrecogniz- independently of the matter itself. A traf-
problem, but missteps abounded. One able after a few turns. fic jam is a familiar type of density wave;
hypothesis was that magnetic fields orga- Lindblad therefore proposed in the while remaining denser than its sur-
nized the patterns, but how they might 1950s that spiral arms are not material roundings, it contains different vehicles
do this hasn’t been explained to most entities but patterns that contain differ- from moment to moment. Sound waves

t=0 t=1 t=2 t=5

Any linear feature in a galaxy’s disk will be wound up into an ever-tighter spiral if the disk undergoes differential rotation, with inner portions
orbiting the galaxy’s core more frequently than outer ones — as is nearly always the case. More than a half century ago this conundrum made
it clear that spiral arms couldn’t contain the same stars throughout a galaxy’s life. Adapted from 21st Century Astronomy.

In all likelihood no one theory can explain all
the structures seen in disk galaxies. Tenth-
magnitude NGC 2613, in Pyxis, has a multi-
tude of short, fragmentary arms like those
spontaneously appearing in the simulated
disks of isolated galaxies.

38 September 2002 Sky & Telescope ©2002 Sky Publishing Corp. All rights reserved.
giant agglomerations of stars, similar to and outside of our Milky Way galaxy.

also are density waves, sequences of mov-
ing air compactions, each more tightly
compressed than the air around it. They
spread swiftly like ripples from their

source, though the individual air mole-
cules within them need barely move.
In a galaxy, the theory holds, density
waves take the shape of a spiral, rotating
independently of its contents. Unlike the

case with sound waves, the driving force
behind density waves is not pressure but
gravity. Stars orbit a galaxy’s center on
their own paths, which rarely cross. But
as a density wave, or spiral arm, ap-
proaches a star, its relatively dense mass Evidence for spiral density waves? The infrared image of NGC 253 (left) favors old, red stars, while
sucks in the star and holds it awhile. The the visible-light image emphasizes dusty molecular clouds and short-lived blue stars. As even
star eventually escapes at the arm’s other the oldest stars in these galaxies’ disks trace spiral arms, the arms must be gravitational “wells”
side to move on until reaching the next that temporarily retard all of the disk’s contents in their orbits around the galaxy’s center.
arm. This process ensures that a star
spends more time in the arms than out a water wave,” says Shu. “Whitecaps are Eternal or Ephemeral?
— perpetuating the spiral and making it always on top because they form and While it held the promise of solving the
stand out in galaxy photographs. break quickly, not because they ride with winding-up problem posed by material
The arms also suck in clouds of inter- the wave.” spiral arms, Lin and Shu’s theory didn’t
stellar gas and squeeze them, prompting Like whitecaps, Shu adds, the visible clearly specify what generated density
some clouds to collapse into dense balls. spiral often looks messier than the un- waves in the first place. Around 1960
These contracting clouds eventually give derlying wave, as it’s much more fragile Lindblad’s son, Per Olof, paradoxically
birth to stars, and the brightest, bluest of amid the galaxy’s turbulence. The young deepened this mystery by conducting the
the newborn stars highlight the spiral stars “are the easiest thing to see of the first simulations of spiral galaxies, pro-
structure. Obligingly, these bright, short- waves, but they’re not the underlying gramming a computer to mimic a galaxy’s
lived stars usually die (as supernovae) structure,” Shu adds. “From infrared ever-moving stars. The onscreen “stars”
before leaving the arms, so they further studies today, we know that the old [and would swiftly arrange themselves into
enhance the visual contrast of a galaxy’s much more numerous] stars in the disk spiral patterns, which would then quickly
spiral pattern. “These young stars are like are indeed the structure’s backbone. You vanish. Even in the computer, it all resist-
the whitecaps, or bubbles, at the crest of can see these arms in infrared light.” ed explanation. But it did raise a ques-

Might the abrupt kink in M100’s eastern
(left) arm be the relic of a disturbing
interaction like that suffered by M51?

M83’s strong spiral arms and weak central bar may
be robbing stars and gas of angular momentum,
slowly driving them in toward the galaxy’s core. ©2002 Sky Publishing Corp. All rights reserved. Sky & Telescope September 2002 39
Are the spirals transient phenomena, destined to unravel after brief glory?

tion: Are the spirals transient phenome- could accomplish this, but they left it to and troughs. Many things can serve as
na, destined to unravel after brief glory? other theorists to examine that question the reflecting barrier. In the boat’s case
Lin and Shu said no, they’re actually more closely. the barrier is the edge of the wake; in a
long-lived — sparking a “short-lived vs. Leaving that aside momentarily, the galaxy, the disk’s outer edge plays that
long-lived” debate that still in some ways Lin-Shu point that ignited the most de- role.
defines research on the spiral question. bate was the prediction that spiral If you look for patterns in this wake,
To understand their argument, imag- patterns should be relatively long-lived. you’ll likely find them in the standing
ine a speeding bullet. It pushes air aside Details may change, it seemed, but basic waves. “Most of the waves will not form
constantly. Each bit of displaced air characteristics — such as the number of standing patterns,” says Shu, now presi-
bumps into the bit of air next to it, creat- arms — persist over much of a galaxy’s dent of Taiwan’s National Tsing Hua
ing an air compaction. The second bit of existence. This is because the pattern University. “They will bounce around
air bumps into the air next to it. The net contains chiefly standing waves. Return- until they disappear.” Only standing
result is a traveling density wave. Waves ing to our speedboat analogy, imagine waves remain clearly evident — in galax-
spread out in sort of a cone around and you’re looking off to the speeding boat’s ies, as spiral arms. Since standing waves
behind the bullet, much as a moving side. Amid the chaos of its wake, some tend to be symmetrical, Lin and Shu feel
speedboat creates a roughly triangular ripples will seem motionless. Stretching this is why many galaxies are fairly sym-
wake. (The two processes are mathemati- in ridges from the front of the boat metrical, two-armed spirals.
cally similar, though water waves aren’t backward, these wavelets are standing
density waves.) waves. Standing waves can arise anytime The Only Constant Is Change
Now let the boat or bullet spin in a lit- waves reflect off a barrier. A reflected One researcher who never totally bought
tle circle, so the waves fan out in a spiral. wave can reinforce one approaching the all this is MIT’s Toomre. He works down
Galactic density waves are exactly analo- barrier, doubling the contrast between the hall from Lin, setting up a close-
gous, the theory says. peaks and valleys. At some places there quarters debate that has simmered for
What originates these waves? Obviously may always be a supply of mutually rein- decades. “I was a young instructor here
there’s no big boat churning the galaxy. forcing waves going both ways. The when Lin came back from an astronomy
Lin and Shu assumed other processes result is a seemingly still pattern of crests meeting in 1961, at Princeton, with this

A Believable Backward Spiral?
hen a team of astronomers announced last January that leading arms, they speculate that an interloper galaxy may
they had found a “backward spiral galaxy,” some re- somehow be responsible. “In the center of [NGC 4622] there’s a
searchers said it could jolt established theories of spiral- dust lane that divides the nucleus almost in half. It could signify
galaxy formation. If the findings are correct, “then spiral-arm that a merger happened in this galaxy,” says Buta.
theory is even farther from explaining reality than we thought,” Two clues suggested that NGC 4622 was spinning the same
says Rutgers University dynamicist Jerry Sellwood (himself not a way the outer arms point. The evidence is much like what you’d
member of the team). But other researchers question the finding. need to determine which way a Frisbee spins. Assuming you see
Standard theories predict that a disk galaxy’s differential ro- the disk at an angle, you would first have to know which edge is
tation would produce trailing spiral arms, whose tips point in a moving toward you and which is moving away. Second, you
direction opposing the galaxy’s rotation — just as if you were would have to know which side is closer to you.
stirring cream into coffee. (The mechanisms are related.) But The researchers got the first piece of information from
NGC 4622 in Centaurus works the other way, according to Gene G. Doppler-shift data acquired with the 4-meter Blanco telescope at
Byrd and Ronald J. Buta (University of Alabama) and Tarsh Free- the Cerro Tololo Inter-American Observatory in Chile. This
man (Bevill State Community College). Its two most prominent straightforward measurement indicated the galaxy’s northern
arms, the outer ones, appear to lead, not trail. “This is probably side is receding from us. The second, more controversial bit of ev-
the only [galaxy] that really shows a leading arm clearly,” says idence came from the fact that dust lanes ring the bright, bulging
researcher Bruce Elmegreen (IBM Watson Research Center), centers of many spiral galaxies. If a spiral’s disk is tilted, dust will
who, like Sellwood, did not participate in the discovery. Byrd, stand out where it crosses the part nearest us, since it is there
Buta, and Freeman announced their findings at last winter’s that it obscures our view of the galaxy’s bulge. Using the Hubble
meeting of the American Astronomical Society and have submit- Space Telescope, the Alabama team found dust obscuration most
ted them for publication in the Astronomical Journal. prominent on NGC 4622’s eastern side, implying that it is closer.
While the researchers have no precise explanation for the What may complicate this analysis, the researchers acknowl-

40 September 2002 Sky & Telescope ©2002 Sky Publishing Corp. All rights reserved.
idea that galaxy disks have these modes,”
or standing-wave patterns, Toomre re-
calls. “It was a decent thought. And it’s
probably true, a fraction of the time —
maybe in 5 to 15 percent of cases.”
Toomre felt most spiral galaxies looked
much messier and less symmetrical than
true standing-wave systems — which
wouldn’t survive long in most galaxies
anyway, he concluded. In the 1960s, he
and student William H. Julian (now at
New Mexico State University) developed Another textbook spiral with its “grand-design” arms, M81 in Ursa Major may be relying on its
a competing proposal that looked more own companion (NGC 3077, not shown here) to periodically pump up its density waves. The
closely at how spirals originate. Very Large Array radio telescope buttressed this notion by mapping the motions of atomic hy-
Suppose a galaxy’s disk contains a rel- drogen gas throughout M81’s disk (false-color image at right). Courtesy Gary L. Stevens and
atively large lump of mass. As this lump NRAO/David Westpfahl (New Mexico Institute of Mining and Technology), respectively.
orbits the galaxy’s core, its gravity at-
tracts other material toward it — though many spiral arms can appear. But, says tic “year” — about 240 million Earth
not directly toward it, as it’s a moving Toomre, they won’t necessarily be inter- years in the case of the Milky Way — the
target. If you sum up the lump’s effects connected or have a coherent “grand de- galaxy’s appearance can change dramati-
on all surrounding parcels of matter, it sign,” as standing waves would. cally. “It’s like an everlasting kaleidoscope
turns out that material bunches into a Toomre believes that in most cases the of shapes,” Toomre remarks. In contrast,
little piece of a spiral arm. That itself is a waves change constantly: the spiral itself Lin and Shu believe the patterns stay
bigger lump, which generates even more jolts stars off the nice, smooth paths that roughly stable for 10 galactic years (2.4
of a spiral arm. Since any galaxy — real feed a good spiral. New spirals constantly billion years in the case of the Milky
or simulated — is “lumpy” throughout, replace old, dying ones. In a given galac- Way) or more.

edge, is that NGC 4622 is tilted only slightly with respect to our
sightline — its disk is no more than 20° from face on. In the
past, the method has been used for galaxies with much greater
tilts, Buta says. But he states that Hubble’s high resolution per-
mits him to extend the technique to galaxies like NGC 4622.
Sellwood, for one, questions whether the method is reliable.
“Whether the dust is in front of, or behind, the stellar disk de-

pends mostly on their relative displacement, not how we view
the galaxy,” he says. Pioneering spiral-structure theorist Alar
Toomre (Massachusetts Institute of Technology) also doubts
that the galaxy is spinning backward. He cites feathery struc-
tures that appear in the second, yellowish arms (counting in
from the outside), which nearly form a ring. Toomre says these
structures resemble classic small-scale manifestations of swing
amplification, a feature of his own spiral-galaxy theory in which
— paradoxically — large trailing spiral
features can arise from much smaller ones
that initially lead. “Any such rudimentary
Almost buried in this Hubble Space Telescope view of 12th-magni-
features simply have to be trailing them- tude NGC 4622 is a dust lane crossing the galaxy’s tight nucleus (bet-
selves, regardless of any minor merging or ter seen in the high-contrast close-up at left). In some astronomers’
interplay that probably did confuse this minds this means that the galaxy spins “backward,” with the tips of
galaxy as a whole.” — J. L. its arms pointing in the direction of the galaxy’s overall rotation.

©2002 Sky Publishing Corp. All rights reserved. Sky & Telescope September 2002 41
The proponents of these theories also admit that the other side might be

Seeding Spiral Structure the nature and persistence of the pattern. the way. In her view, stars across most of
Although these competing theories re- The proponents of these theories also a galaxy’s disk give angular momentum to
main at odds, rough agreement has admit that the other side might be right the spiral wave. As a result, they sink ever
emerged on forces that could initiate spi- — in some cases. But otherwise they so slightly toward the center each time
ral waves. Many researchers believe that have deadlocked in recent years, opening they cross a wave crest. The wave carries
almost any disturbance of a churning the way to new theories. the angular momentum to the outer
galactic disk can launch them. A spiral One of these novel alternatives contains disk, whose stars consequently drift even
galaxy is born of a prototypical gas cloud some unusual ideas that have won praise farther out. Eventually, the spiral galaxy’s
that flattens under its own gravity into a from some astronomers while leaving inner disk contracts into a featureless
pancakelike disk. The disk’s contents re- others befuddled. Yet it started with a lump more akin to an elliptical galaxy.
volve, thanks to spinning motions in the simple notion. Now a senior scientist The theory, Zhang declares, explains
original cloud, and the resulting cen- with the Naval Research Laboratory, Xi- many otherwise surprising observations.
trifugal forces keep it from collapsing into aolei Zhang was a graduate student at For instance, recent surveys by the Hubble
a dense sphere. A wave-making distur- the University of California, Berkeley, in Space Telescope and ground-based tele-
bance can come from lumps in the disk, the late 1980s when she was struck by scopes have found that a greater propor-
which jolt the system by creating an in- the image of a spiral galaxy in the as- tion of galaxies are ellipticals today than
ordinate gravitational tug in their neigh- tronomy department’s hallway. “My first was the case in the far past, whereas the
borhood. It can also come from outside, impression was that stuff must be spiral- total number of galaxies appears roughly
from a nearby galaxy’s tidal pull; Toomre ing inward,” she says. A professor, though, the same. The opportunity may come to
believes this explains symmetrical galax- told her that this is an illusion, because test Zhang’s notion when the Next Gen-
ies like M51. Also, many galaxies contain the stars’ orbital angular momentum eration Space Telescope replaces Hubble
barlike spinning structures, which, as- keeps them from falling into the center. in 2010. This will enable astronomers to
tronomers believe, can exert the gravita- Undaunted, Zhang started looking for scrutinize spiral-galaxy evolution in much
tional torque (twisting force) required to ways that stars might lose angular mo- greater detail.
excite waves. mentum and thus spiral in.
Both competing theories also describe Several years later, in three Astrophysi- Order and Chaos
how small initial disturbances might grow cal Journal papers, Zhang — then at the In a larger sense, Zhang says, galactic spi-
into the large-scale patterns we see. But Harvard-Smithsonian Center for Astro- rals are related to a tendency — which
they disagree on the mechanism, and on physics — argued that the spiral itself is scientists have long recognized — of all

“We argue that most visible spirals are short-
lived features organized by gravitational forces,”
wrote Jerry Sellwood (now at Rutgers University)
and Ray Carlberg (University of Toronto) in 1984.
Their conclusion came from pioneering computer
simulations in which 20,000 “stars” in a flattened
disk attracted one another while also respond-
ing to the gravitational effects of a dark-matter
halo enveloping the entire galaxy. Each frame
spans 65,000 light-years. This updated version,
with 10 times as many particles, was computed
for Sky & Telescope by Jerry Sellwood.

Like Sellwood, Xiaolei Zhang (Naval Research
Laboratory) has found that spiral structure can
arise spontaneously in a rotating disk of stars.
But rather than emphasizing the structure’s tran-
sient nature, her models exhibit some long-lived
features, like the two dominant spiral arms in
the inner half of the disk shown here. Taken to
represent a galaxy like our Milky Way, these
frames are 140,000 light-years wide. Courtesy
Xiaolei Zhang.

42 September 2002 Sky & Telescope ©2002 Sky Publishing Corp. All rights reserved.
right — in some cases.

physical systems to increase in entropy,
or disorder. Paradoxically, a spiral helps a
galaxy reach disorder, though it itself is
ordered. The concept stems from the con-
troversial proposals of a Belgian Nobel
laureate in chemistry, Ilya Prigogine.
Now a physics professor at the University
of Texas, Prigogine has argued that many
physical systems far from equilibrium
sometimes try to reach that state of max-
imum disorder by forming a coherent Start 160 million years
structure. For example, a sharp tempera-

Heikki Salo (University of Oulu, Finland) and
fellow researchers have programmed com-
puters to see what happens to the stars in
one galaxy’s rotating disk when a second gal-
axy passes it at close range. As shown in this
model of the M51–NGC 5195 system in Canes
Venatici, tidal extensions get wrapped into
spiral arms by the disk’s differential rotation.
While the Whirlpool’s famous spirals often
are attributed to its interaction with smaller
NGC 5195, models like this allow for the pos-
sibility that the Whirlpool already had some
spiral structure before the encounter. A cross
represents smaller NGC 5195, and the ellipti-
cal arcs show its presumed orbit around M51. 320 million years 540 million years
Courtesy Heikki Salo.

©2002 Sky Publishing Corp. All rights reserved. Sky & Telescope September 2002 43
Shu predicts that spiral research may lead into uncharted scientific territory.

ture difference between the upper and ies are assumed to contain stars with a findings suggest an uneven angular-
lower atmosphere will often generate or- wide range of angular momenta in momentum distribution in our region of
ganized patterns of convective air cur- roughly equal quantities — that is, a fair- the galaxy, says Sellwood. However, he
rents. These swiftly mix up the air, re- ly uniform distribution. But this is un- admits, Hipparcos didn’t collect some
ducing temperature differences. Zhang proven. In fact, Sellwood says, a galaxy data needed to confirm his theory. Sell-
says that galactic spirals organize them- could have one or more “grooves” in its wood thus is awaiting the much more
selves in a similar fashion. angular-momentum distribution — a copious and precise astrometric data that
The notion is faintly disturbing. It shortage of stars with particular amounts ESA’s Gaia spacecraft should produce
seems to almost suggest that inanimate of angular momentum. In computer after its launch a decade from now.
things “want” to act a certain way. Yet simulations, a groove will “destabilize the For his part, Shu predicts that spiral
Zhang’s theory is by no means the first disk and produce a spiral pattern,” Sell- research may lead into uncharted scien-
to do so. Scientists sometimes say things wood says. The spiral itself produces two tific territory. Some parts of a spiral gal-
fall because they “want” to become more more grooves, because in certain places axy, he observes, revolve in one direction,
tightly bound in a gravitational field; the frequency (or rhythm) of the spiral’s while others orbit the other way. If you
similarly, stars in a galaxy “want” to rotation resonates with those of nearby add up the angular momenta throughout
move closer to the system’s center. Most stars’ orbits. This scatters stars about, just the pattern, Shu says, you may get zero.
theorists agree that spirals help achieve as a wave can give a great push to a This implies that spiral patterns might
this, says Rutgers University’s Sellwood. surfer moving at the same speed as it pull themselves up by their bootstraps,
On a different tangent, Sellwood re- does. The new grooves create more spi- from nothing — a type of process sci-
cently developed a hypothesis that puts a rals. This creates complex, ever-renewing ence has barely begun to grapple with.
new spin on Toomre’s proposals. Sell- patterns, which Sellwood calls “recurrent The universe itself may have begun simi-
wood believes that lumpiness alone might dynamical instabilities” and likens to a larly, Shu adds.
not suffice to launch large-scale waves. flag flapping in a breeze. “Sometimes,” says Shu, “you can get a
He proposes a different process related to Sellwood says that his view has tenta- free lunch.”
stars’ angular momentum — the product tively been confirmed by the European
of each star’s orbital velocity around a Space Agency’s Hipparcos satellite, which Science writer Jack Lucentini (jekluc@aol
galaxy’s center and its distance from the measured stellar distances and motions .com) covers astronomy, space science, and
center. Normally, Sellwood explains, galax- in the Sun’s neighborhood. Hipparcos computing from his home office in New York.

If we could see our own Milky Way galaxy from outside, we might see a large number of thin, relatively tightly wound arms and a weak central
bar. This illustration is based on decades of observations, from radio-telescope maps of atomic hydrogen gas to microlensing surveys of the gal-
axy’s bulge. Reprinted from 21st Century Astronomy by Jeff Hester, David Burnstein, George Blumenthal, Ronald Greeley, Bradford Smith,
Howard Voss, and Gary Wegner. With permission of the publisher, W. W. Norton & Company, Inc.

Galactic Bulge 3 kpc Arm
Norma Arm
Scutum Arm Crux Arm

Carina Arm
Perseus Arm

Orion Arm
Outer Arm Sun
Sagittarius Arm


44 September 2002 Sky & Telescope ©2002 Sky Publishing Corp. All rights reserved.