Classical cosmology was once described as ``a search for two numbers'': the rate of cosmic expansion and the parameter that betrays the universe's fate (perpetual expansion or eventual collapse). Thankfully, cosmology is far more interesting today. We have dark matter, temperature fluctuations in the microwave background, Great Attractors, Great Walls and perhaps something so new that it's been announced since I wrote this article.
It used to be that astronomers were those guys that worked in the cold with stop-watches that didn't need second hands. They heroically guided telescopes to keep hour-long exposures from blurring. Then came all those chirping high energy sources of the sixties. It made it all so much more exciting. But, surely cosmologists just work on fastastically faint things at the edge of the Universe. Once again, they must be tied to telescopes night after night for a single frame.
Sure, those guys are around. We don't need many of them to burn major telescope time. But, why make cosmology so faint and hard? Think of the Local Group of galaxies. There are a couple dozen probes of the dynamics of the Universe. Some of them race toward us while others still expand away. All of the galaxies are old, they must have formed and evolved. The entire group must have formed somehow. So, let's look at the cosmology of nearby galaxies.
The number of galaxies in the Local Group is slightly uncertain. We don't know all their distances and the cutoff for membership is ambigous, but there are currently about 30. All of the probable members are listed in Table 1 together with best values for their distances, apparent magnitudes and sizes, where best involves a wide range of quality. The more distant and ratty dwarf galaxies are hardest to pin down. When you go to the rattiest of them all, little snippets like Ursa Minor, we haven't even done very well at measuring their apparent magnitudes. While at least 1/10 of an iceberg sticks out above water, the maximum surface brightness of these dwarfs is far less than the night sky. Similar apologies apply to the diameters listed. They are the scale on the main structure on a deep photograph, not a well-defined uniform set of measurements. This bunch of galaxies are anything but ``well-defined'' and ``uniform'', so that's too much to ever hope for. Figure 1 shows the Local Group without the outlying members.
Table 1
The biggest and brightest Local Group members are the Milky Way Galaxy and the brightest Messier objects: M31 and M33. Next in line would be M32 and the two Magellanic Clouds (the LMC and SMC). The Clouds are big and close, so we have good detailed studies of them. They have even been detected in North America-if you count the neutrinos from Supernova 1987a. The rest are smaller objects, either irregular galaxies or dwarf ellipticals.
Two other more distant and less luminous irregular galaxies that have gotten well deserved beatings with large telescopes are NGC 6822 (``Barnard's Galaxy'') and IC 1613. They are providing new insights for both the distance scale and the evolution of galaxies. Both have Cepheid variables, still the best way of determining distances within the nearest 10 million light-years, and both have current star-formation activity. For reasons as yet unknown, star formation in NGC 6822 is far more active than the otherwise similar IC 1613.
In the basement of luminosity, we find the ``Seven Dwarfs'', the small, very faint dwarf elliptical galaxies that surround the Milky Way Galaxy. They are neither spirals nor irregulars. However, they are so different from any other elliptical galaxies that we will follow convention and call them Dwarf Spheroidals or DSph's. There might be a lot of these dwarfs around other galaxies, but they have been hard enough to find nearby. A recent exciting result has been the discovery that these objects are completely dominated by dark matter, as we will discuss later.
This rat pack of DSph's is now up to nine with two more having been found in 1990. The Sextans DSph was discovered by M. J. Irwin and his collaborators while searching for quasars with Cambridge University's automatic plate scanner. The surface brightness of the Sextans DSph is so low that your eye can't find it on the discovery plates. It was only uncovered by noticing a slight excess in the number of faint stars in its direction. It doesn't look much better on deep CCD images. But, Mario Mateo and collaborators at the Carnegie Observatories were able to use such images to get a color-magnitude diagram, determine a distance of 290,000 light-years and an apparent visual magnitude of 10.3 (that took some heavy handed assumptions regarding the ratio of observed stars to the total number). However, don't go out and look for a tenth magnitude galaxy in your finder; the luminosity is smeared out over an area greater than 100 times the size of the moon.
Gary Da Costa and collaborators at the Anglo-Australian Observatory took spectra of six stars in the Sextans DSph and found that its moving rapidly at 230 kilometers per second. While they found that the stars are low in heavy elements, there are far more heavies than in the comparable dwarfs, Ursa Minor and Draco. Perhaps this will prove to be a clue to the mystery of how the dwarf elliptical galaxies formed and evolved.
Russell Lavery of the Mt. Stromlo Observatory gets credit for the second extreme dwarf discovered in 1990. His dwarf lies in the southern constellation of Tucana and he called it a nearby DSph, but others are claiming that it's a more distant dwarf irregular. In any case, it has been resolved into stars, so it's almost certainly a member of the Local Group.
What is the criterion for inclusion in the Local Group? Proximity is the cleanest and often used to the exclusion of any other. If we grant membership to all galaxies within 4 million light-years, we have a club with 30 members, three of which barely got in. We can also use velocities to find out if the last three are on their way in or out. That is, we can accept all the fellow travelers. If we do this, the three squeakers become full members and we are pressed to include a few more distant objects such as Leo A and Pegasus, both small irregular galaxies.
There are still other applicants on the waiting list. There are two faint irregular galaxies in Sextans, imaginatively designated Sextans A and B, both of which are well-resolved into stars. At the moment, they just barely fail to qualify. The bright irregular galaxy NGC 3109 also hovers on the periphery with a velocity that is only a little too large. The membership committee will have to carefully consider another small irregular object nominated by R. Kraan-Korteweg and G. Tammann, UGC-A86. A recent letter of recommendation has been published by A. Saha and J. Hoessel. Their supporting documents include a color-magnitude diagram that fixes a distance close enough for inclusion. The applications of past candidates, such as the heavily-obscured spirals NGC 6946 and IC 342, and the Maffei 1 and 2 infrared galaxies, have been declined. They fail on all counts, their velocities are all wrong and they are too distant.
We've talked about ellipticals, spirals and irregulars-classifiers used
for normal galaxies. Are there no active galaxies in the Local Group?
Well, remember when psycology used to be divided into abnormal and normal
disciplines? It seems the 1970's and 80's brought so many neuroses and
situational traumas, that those designations fell by the wayside. The
normal/active galaxy dichotomy seems destined for a similar fate. While
there are no quasars (the ultimate galaxian psycotics) in the Local Group,
but we do have our share of neurotics and traumas. The center of our
galaxy mysteriously spews out a jet of positrons (the antimatter mate of
electrons) that seems to turn on and off. Black hole hunters have found
their beasts lurking in the cores of M31 and M31 (August issue, page 142).
The orbit of the Magellanic
clouds brings it so close to the Milky Way that tidal shocking has led to
episodes of discrete star formation and the tearing off of a tidal tail.
This Magellanic Stream is a ribbon of gas that can be traced over nearly
. It defines a polar ring, plane that is perdendicular to the disk of
the galaxy. Mysteriously, several of the DSph's lie in this plane.
DARK GALAXY DYNAMICS
The Local Group offers a unique laboratory to compare the distributions of light and mass. The display box (Measuring Masses) explains that the key to discovering dark matter is determining distances and velocities. Let's turn to these measurements in the Local Group. (A short version of this box is also shown as Figure 2).
In 1912, Vesto Slipher of Arizona first discovered that M31 was racing toward us with a velocity of 300 kilometers per second. He also detected a ``tilt'' in the spectral lines that owed to the rotation of the galaxy. By 1917, F. G. Pease had used the 60-inch at Mount Wilson to measure the velocity as a function of radius, the first rotation curve of a galaxy ever produced. Five years later, Ernst Öpik at Tashkent used this rotation curve to find that M31 had 2-3 times as much mass for its light compared to the Milky Way. However, rather than conclude that it was a ``dark galaxy'', he proposed revising the distance. This confusion between darkness and distance arises because when we used observed quantities like angular diameter and velocity to measure a mass, we find that the mass we infer scales as the distance to the galaxy. However, when we go to determine the luminosity, we use the locally measured flux times the square of the distance. So, if we mistakenly measure a distance that is half of the true value, we get a luminosity that is 1/4 of the true value, but our mistaken mass is still 1/2 of the right answer. When we take the ratio to get a mass-to-light ratio, M/L, its 2 times the true value. This may lead us to think that the galaxy is dark. Öpik results on M31 caused him to suggest that it was twice as far away as Hubble claimed at the time and even this falls short of the modern distance.
A decade later at the Sterrenwacht in Leiden, Jan Oort compared an
inventory of all the known stuff near the sun with the mass density
determined by the motions of stars perpendicular to the plane. Oort's work
provided a benchmark value of M/L. He found 
for
the observed mix of stars, but the inventory only accounted for a little
more than half of the mass measured dynamically. Both the dynamical
measurment and the inventory were a tricky business. Oort's result has
been continually reexamined and, until recently, it was always confirmed.
However, astronomy is full of cautionary tales regarding the comparison of
two tricky measurements that differ by a factor of 2. Recently, Gerry
Gilmore of Cambridge University has argued that the inventory was 
low,
the mass had been 
high and the errors on both are 
. So, you can
neither insist that there is unexplained mass locally nor can you rule it
out. John Bahcall and his group at the Institute for Advanced Study still
maintain that 
1/3 of the total must be dark. This jury of one is still
out.
Just a year after Oort's work, Caltech's incomparable Fritz Zwicky used the
random velocities of galaxies in the Coma cluster to measure it's mass.
When he compared this to the luminosity, he got a stunning 
.
Zwicky's 500-fold increase in the expected mass would seem far more
compelling that Oort's missing half, but remember Öpik. The current
distance to Coma is one-tenth of the value that Zwicky used. Coma is still
at the foundation of the case for dark matter, but with the less extreme
value, 
.
Within the Local Group, there are techniques for measuring distances that we are only beginning to be able to exploit on larger scales. This removes much of the ambiguity between distance and darkness. We can probe the dynamics of the Local Group in unique ways and study the dwarf spheroidals which are only visible within the Local Group.
These advantages have been continually exploited. After Slipher discovered
that M31 was approaching the Milky Way back in 1912, he found that 36 of
the next 40 galaxies moved away. In the late 1920's, Howard Robertson
(Caltech) and Edwin Hubble (Mount Wilson Observatory) working independently
north and south of Rose Bowl parade route, both concluded that the Universe
was expanding. In 1959, while they were both in Princeton, Frank Kahn and
Lodewijk Woltjer realized that something must have reversed the expansion
of M31. The only reasonable cause is the mutual gravity of the dominant
galaxies. Knowing the velocity of approach and the age leads to a measure
of the mass. The result was a stunning
or an M/L in excess of 30 M/L. There must
be 10-20 times as much nonluminous matter as luminous!
This result is corroborated by a number of measurement techniques that are
unique to the Local Group. Distant globular clusters and small satellite
galaxies can be used to measure the mass of the Millky Way. Requiring each
one to be bound leads to independent estimates of the mass of the Milky Way
out to their radii. The inferred mass climbs as we look at more distant
dwarfs, reaching 
in the case of Leo I (determined by Dennis
Zaritsky and collaborators at Steward Observatory). The Magellanic clouds
offer an extra bonus. The spatial structure and kinematics of the
Magellanic Stream are most easily understood if it was torn off the clouds
at the time of their last closet approach. Independent groups in Japan and
England used the velocity of the tip of the stream to find that the mass of
the Milky Way must be . Final corroboration for this large mass comes
from the high velocity of the new distant DSph in Sextans.
The velocities of every other member of the Local Group offer challenging tests for the picture of two massive galaxies formed by gravity in an expanding Universe. Last year, Jim Peebles (Princeton) found that the model passes every one of these tests.
Peebles has also stressed the remarkable cosmological consequences of the dynamics of the Local Group. In recent years, the Inflationary model of the Universe has achieved a high degree of success. In this model, the Universe was flattened by a phase of rapid expansion during the first 1E-39 seconds of the Big Bang. The result should be an exquisitely balanced Universe whose density causes it to teeter on the precipice between being bound and unbound. It will expand forever, but any overdensity-no matter how tiny-will ultimately collapse. However, the models of the local group show that the Universe must have a density that is only 1/10 of the critical value required by the Inflationary cosmology. The collapse of the Local Group is causing the conventional picture of the first 1E-39 seconds of the Universe to implode!
HEARTS OF DARKNESS
Over the last two decades, we have discovered that as much as 
of the
Universe is dark and we don't know what it is. Most of our insight into
the ratio of light and dark matter has been gained from the Milky Way and
the Local Group. There are lots of questions about dark matter that could
help us sharpen our search for the magic stuff. One promising diagnostic is
the behavior of darkness as a function of galaxy brightness-and the local
group is the best place to stretch that to the faint end.
Big galaxies like the Milky Way are bright in the middle and dark around the edges. Presumably, the gas that eventually made the stars sunk into the core of the dark matter because it was able to lose energy and the dark matter couldn't do that. If the dark matter was once ordinary gas, it must have turned into underluminous stars quickly before it had time to sink into the center. A wildly popular idea is that the dark matter is some sort of exotic matter that was unable to lose energy and is forced to remain in an extended halo. In 1982, Marc Aaronson of Steward Observatory discovered that the dwarf spheroidal galaxies, Draco and Ursa Minor, have hearts of darkness-their central M/L's are nearly 100 M/L.
These dwarfs have had hearts of stone when it comes to theories of dark matter. Aaronson pointed out that these galaxies were too small to have captured any ``hot dark matter'' like neutrinos. Recently, I've modeled these dwarfs and found that the density of the dark matter in their cores is ten times the central value in the Milky Way-large enough to heap more dirt on the grave of ``cold dark matter''. A remarkable feature of the dark matter in the dwarf spheroidals is that while the central column densities of stars are 1/1,000 of that seen in bright galaxies, the central column density of dark matter is the same. These features argue that some, if not all, of the dark matter is in the form of low mass stars. If that's true, the dwarf spheroidals will have a faint infrared glow visible with the next generation of IR space telescopes.
WEIRD GRAVITY
Rotation curves are an extremely important part of dark matter research, yet we've scarcely mentioned them here. In 1959, the radio dish at Westerbork was used to get the first radio rotation curve of a galaxy-M31, of course. Just 5 years before that work, Martin Schwarzschild wrote a magnificent paper where he calculated a rotation curve for M31 based on its luminosity profile and compared it with existing optical data. The agreement at the time was good, but Schwarzschild noted that there could be dramatic changes in stellar populations with radius and clearly stated the importance of getting extended rotation curves. When the first radio rotation curve for M31 was published by Dutch astronomers, there was a companion paper with a mass model built by Maarten Schmidt. Curiously, there was no tabular or graphical comparison of the data with either Schmidt's or Schwarzschild's model.
In 1963, Arrigo Finzi of Rome finally made that comparison and found that it was lousy-something was going on at large distances. He was the first to collect all of the evidence that we have described for dark matter and concluded that the law of gravity must change with distance. His argument was simple: when the force of gravity is very strong, Newtonian theory fails and we must use General Relativity. Similarly, when the force of gravity becomes weak in the outer reaches of galaxies and clusters, even General Relativity might fail. In the last decade, Mordehai Milgrom of the Weizmann Institute has suggested a similar revision of gravity at low acceleration rather than large distances. In these theories, dark matter is as illusory as the planet Vulcan.
The test of the Finzi/Milgrom ideas is simple. In their theory, the dark matter problem is part of a Universal Law of Motion. All galaxies that look the same and are in the same environment must appear to have the same amount of dark matter. In the standard dark matter picture, the problem is regrettably more like bartending than physics. Just as we can mix a Martini wet or dry, we can conceive of galaxies mixed dark or not. The dwarf galaxies are indeed mixed both ways. IC1613 and Fornax have little, if any, need of dark matter, whereas Draco, Ursa Minor and Carina are dark at their very cores.
The richness of the Local Group continues to emerge. We can test ideas about the first 1E-39 seconds of the Universe, test the law of gravity and increase our understanding of dark matter-all by looking at the our big sibling Andromeda and those lovable Seven (now nine) Dwarfs.
