by Bill Keel

Introduction - Island Universes

Galaxies are the lighthouses that plumb the Universe - constituents of the largest-scale texture we know. They span a vast range of properties, from dwarf galaxies with a few million stars barely outshining the  brightest individual star clusters in our own galaxy, to vast assemblages of a trillion stars in the centers of great clusters. Our own galaxy, a reasonably bright spiral system, can be traced at least fifty thousand light-years from its nucleus, and we know of many galaxies much larger  yet. Some elliptical galaxies show no evidence of having formed stars since a brilliant epoch early in cosmic history, while spiral and irregular  galaxies have been making stars briskly over their entire lifetimes. Some  galaxies produce most of their energy deep in the infrared, and some are  so diffuse and faint as to be barely detectable against the faint glow of the Earth's night sky.

Some History

Our appreciation of the universe beyond the Milky Way is entirely an achievement of the twentieth century. The objects which we now know to be galaxies had, to be sure, occasionally drawn the curiosity of visual observers from the days of Charles Messier forward, particularly William Parsons (the Earl of Rosse), whose 72-inch (1.8-meter) telescope with its speculum-metal mirror had revealed the intriguing spiral forms of certain dim, cloudy objects (nebulae) seen, by and large, far from the encircling band of the Milky Way. However, further tools were to be needed to unravel the true nature of some of these objects.

By the 1920s, photography had revealed that there must be tens of thousands of these objects, by then known as white nebulae to distinguish them from the clearly different gaseous nebulae such as the famous Orion Nebula, accessible to the telescopes of the time. They showed a variety of spiral, elongated, or oval forms. The most plausible theories to account for these nebulae made them either nearby objects - perhaps planetary  systems in formation—or extremely distant, truly ``island universes" of which our Milky Way, hitherto the entire known Universe, would be merely one among myriads.

The key observation in resolving this dispute came from Edwin Hubble, using the recently completed 100-inch (2.5-meter) telescope on Mount Wilson, California. Targeting the largest and brightest of the ``white nebulae", as the ones most likely to be nearby in space, he repeatedly photographed selected portions of them as deeply as the available photographic plates would allow. Faint starlike points had been recognized in these nebulae, but could one show that these were in fact stars such as we know in the solar neighborhood, and thus at the enormous distances required to make them appear so faint?

[I put English units first in the above paragraphs because the designations of these telescopes incorporated their apertures in inches; where this is not the case I put metric first.]

Hubble's breathrough came in identifying stars with a particular kind of cyclic change in brightness, which them ``standard candles" whose absolute brightness could be determined - Cepheid variable stars. Henrietta Leavitt at Harvard had shown that this class of pulsating stars has the useful property of a tight correlation between the period required to complete one pulsation in surface temperature and size (and thus brightness) and the amount of energy the star gives off (usually expressed as absolute magnitude, the stellar brightness which we would measure if a star were located at a reference distance of ten parsecs). Cepheid variables gave Hubble the first necessary yardstick in the ladder of extragalactic distances. (One of the major programs for the Hubble Space Telescope is the measurement of galaxy distances beyond the reach of ground-based instruments, by identifying Cepheid variables in more and more distant galaxies. One HST team has recently reported success in measuring Cepheids in galaxies of the Virgo Cluster, about twenty times as distant as the galaxies Hubble the astronomer was observing).

This discovery, in one stroke, opened a whole new vista of the Universe. Within a decade, many of the major strands of galaxy research had begun. Clusters and groups were recognized, classification schemes were proposed, and spectroscopic measurements were begun. Spectra of galaxies proved especially rewarding. Early measures by V.M. Slipher at Lowell Observatory, using very delicate multi-night exposures, had shown that some ``spiral nebulae" exhibited unusually large Doppler shifts. It eventualy developed that galaxies exhibit, on average, a relation between the redshift of features in their spectra and their estimated distances. This gave a way to estimate the distances of ever fainter and more remote galaxies, and provided the first glimpse of an expanding universe.

Kinds of Galaxies

Scientists and naturalists alike have the urge to sort, classify and organize new phenomena, in the hope of seeing underlying patterns that have physical meaning. Several classifications for galaxies were proposed early in their study; Hubble's classification system has proven remarkably robust, correlating well with physically interesting measurements such as stellar content, gas content, and environment despite being designed only to describe the appearance of the galaxy as seen on photographs with blue-sensitive emulsions. With later extensions by Gerard de Vaucouleurs and Sidney van den Bergh, this remains the most commonly used description of galaxy forms.

Elliptical galaxies were denoted by the letter E and a number describing the galaxy's apparent shape - 0 for a completely round form, 5 for one twice as long as wide, and 7 for the apparently flattest genuine ellipticals. We do not know, solely from an image, the true shape of such a galaxy; the same galaxy might have quite different degrees of flattening if viewed from different directions. Elliptical galaxies are, in general, characterized by old stellar populations and very little of the gas and dust needed to form new stars.

Spirals are divided into ordinary and barred spirals; in barred systems the spiral arms arise from a straight "bar" passing through the center, while ordinary spirals have a more S-shaped inner configuration. Ordinary spiral are denoted S and barred systems SB. Both usually contain a central bulge, often sharing many properties with elliptical galaxies, surrounded by a thin rotating disk containing whatever spiral structure there may be. Spirals are subdivided into a sequence jointly defined by the winding and prominence of the spiral arms, and the relative importance of the central bulge. Sa galaxies have a bright bulge and tightly wound arms, while Sc galaxies have loosely wound arms and a relatively less important bulge. This sequence Sa-Sb-Sc-Sd has counterparts SBa-SBb-SBc-SBd in the barred spirals. As more detail was observed in some galaxies, intermediate substeps (Sab,Sbc,Scd,S0/a) could be added when necessary.

Some galaxies show no particular organization, either because some recent event has left them in a disturbed state or because they simply lack the organizing rotation and wave motions of a spiral. These are simply called irregulars; the ones that do not result from external disturbance form, in many respects, an extension beyond Sd of the spiral sequence.

Hubble recognized that the connections among various types left an apparent hole which he called S0 - objects with a bulge and disk, but little or no star formation, dust, or gas. They would form a bridge between ellipticals and spirals. Later, many genuine S0 galaxies were in act recognized, and understanding their origin promises to tell us much about the history and development of galaxies in general.

Several refinements of the Hubble classification have proven widely useful. Gerard de Vaucouleurs introduced distinctions depending on whether the spiral structure proceeds from the nucleus in an S-shape (s) or from an inner ring (r), or some combination (rs) or (sr). He also recognized intermediate cases SAB between barred and nonbarred galaxies. These new dimensions allowed a finer discrimination of galaxy structure and opened the way for more detailed study of the physical properties of spiral galaxies. Sidney van den Bergh noted that the most luminous spirals have long, well-developed spiral arms, and introduced a luminosity classification driven by the organization and distinctness of the arms; Sc I galaxies are in the mean the brightest Sc galaxies, and Sc V the faintest. Note that the classification is based solely on a galaxy's appearance, with its absolute magnitude a correlating quantity.

Some galaxies are not well described by the Hubble system or its variants, even excluding "train wrecks" resulting from galaxy collisions. There exist, usually in rich clusters, enormous elliptical-like systems that may span millions of light-years with more extended outer regions than a similarly huge elliptical would show. These are given the designation cD, from a scheme developed at Yerkes Observatory by W.W. Morgan. Dwarf galaxies may be irregular, elliptical, or spheroidal, depending on their degree of symmetry and central concentration. Recent work has turned up galaxies of very low surface brightness, which must have had a rather different history from familiar spirals. While many of these look like the ghosts of ordinary spirals, it is not at all clear how they connect to the familiar Hubble types.

Content of Galaxies

We observe stars, gas, and dust in galaxies. Stars come in a wide range of age and mass, and are intricately linked to interstellar matter by processes of stellar birth and death. This means that galaxies have a history, which we can probe either by investigating the makeup of a galaxy in detail, or in a kind of fossil probe unique to astronomy, look at galaxies so distant that the light we observe left them when they were much younger than they are "today".

In tracing the makeup of galaxies, there are numerous clues as to the populations of stars present. Different kinds of stars (giant/dwarf, hot/cool, higher/lower abundances of heavy elements) have different patterns and intensities of features in their spectra. In most galaxies, we can observe only their overall (integrated) spectrum, so that a mathematical solution can give constraints on the overall population, but the solution is not completely well-determined without additional assumptions (such a a smoothly varying star-formation rate, or fixed ratios of stars at various masses). To resolve these ambiguities, observations of very nearby galaxies are crucial, where individual (luminous) stars can be observed and counted.

Some components of a galaxy stand out in specific kinds of observations, so that interstellar matter and certain kinds of stars can be studied in isolation. The 21-cm radio emission of cold atomic hydrogen traces this component of a galaxy cleanly, giving one index of its gas content and tracing internal motions beautifully. The gas most immediately associated with the birth of new stars is colder and denser than neutral hydrogen, being mostly molecular hydrogen and best observed via the trace molecule CO, which emits spectral lines in the 1-3 mm range. The most massive young stars emit copious ultraviolet radiation, which may be absorbed by surrounding gas and re-emitted as spectral lines including H-alpha in the visible region, so that using special filters and image processing allows a view of these star-forming regions alone (so long as they are not hidden from view by intervening dust). The dust itself emits longer-wavelength infrared radiation, so we can trace the location of interstellar dust and locate the regions where it is warmed by starlight. Going to the ultraviolet, only the hottest stars give off enough radiation to see, so this region also allows us to trace regions of active star formation. Finally, loking at a galaxy in X-rays, we see only the highest-energy components - binary stars in which material falling onto a neutron star or black hole gives rise to extreme temperatures, emission from gas at millions of degrees, and sometimes emission from central quasar-like active nuclei which may not give an accurate indicator of temperature, since so-called nonthermal processes may be involved.

There is growing evidence that we may be completely ignorant of one of the most important constituents of galaxies - the dark matter. If gravity behaves over ranges of thousands of light years in the way that it does over smaller scales, the motions of stars and gas in galaxies, and of gas and galaxies in clusters, require that most of the mass in these systems is on some completely invisible forms. The main lines of evidence include:

The nature of this unseen matter remains elusive, and has provided a happy hunting ground for observer and theoretician alike. Proposals have included brown dwarf stars, Jupiter-like objects, quantum black holes produced in the early Universe, and a whole zoo of exotic particles which would also be remnants of the early Universe. Assorted astronomical and laboratory searches have yet to tell us what makes up most of the matter in the Universe. We are left with the sobering realization that all of our vaunted technology and apparatus has been telling us about only 10% of the cosmos.

Clusters of galaxies

Early surveys of galaxies on the sky showed that certain regions have more than their share of galaxies; such concentrations as the Virgo cluster were known long before the nature of galaxies was understood. More complete statistics have shown that the distribution of galaxies in space is far from the uniform "sea" first envisaged, with many (perhaps most) galaxies arrayed in groups, clusters with thousands of members, superclusters, and even larger sheets and fingers stretching as far across the Universe as we can reliably map.

Clusters come in a variety of kinds, just as galaxies do. The richest and densest clusters are round assemblages, while sparser clusters have flattened or irregular shapes. The cluster environment is reflected in it sgalaxy content - dense environments like cluster cores are populated almost solely by elliptical and S0 galaxies, nearly devoid of gas and star formation. Less extreme environments can host, as well, spiral and irregular galaxies. This so-called morphology-density relation has engendered a classic heredity-environment question - were spiral galaxies never formed in those regions which would one day be rich clusters, or are they somehow destroyed or transformed in such clusters? The jury is still out, though there is strong evidence that in some clusters spirals were once numerous and have been tranformed by external factors into elliptical or S0 systems. One such transforming mechanism is via galaxy mergers, which, while not common at the high speeds typical of cluster encounters today, might have been more common early on.

A second transforming mechanism could be provided if clusters contain some kind of external medium - intergalactic gas. Such a medium was indeed discovered by early X-ray astronomy satellites, and is known to be ubiquitous in clusters and even galaxy groups. Random motions in the cluster heat this gas to temperatures of 10,000,000 Kelvin, [alas, SI usage is that there is no such thing as degrees K] making it visible only by its own X-ray emission. This gas typically has as much mass as do stars in the visible galaxies, and as galaxies move through it, will provide an external wind. This would in principle be strong enough to sweep gas out of a spiral galaxy, and a gas-free spiral will cease star formation and quickly look like an S0. Detailed observations in local clusters such as Virgo in fact show that spirals nearest the center seem to have lost the outermost parts of their gas distributions.

Detailed studies of the distance and redshifts of nearby galaxies have added another dynamic aspect to our understanding of clusters - they are still growing. At greater and greater distances, the gravity of a cluster takes longer to affect the motions of its surrounding galaxies, so that galaxies at larger distances will eventually turn around against the expansion of the Universe and fall into the cluster. Our own local group has a detectable motion toward the Virgo Cluster (more precisely, the core of the Local Supercluster), and such large-scale motions can be found near many nearby clusters. In this sense, the cosmic epoch of cluster formation is now.

Galaxies and cosmology

The very recognition of galaxies as objects at vast distances led to the first attempts to use thems as tracers of the structure of the Universe as a whole - observational cosmology. Hubble's promulgation of the evidence for a relation between a galaxy's distance and its redshift led to a picture of an expanding Universe. As observational capabilities have increased, so has the volume of space where astronomers can search for signatures of the geometry of space-time.

The Hubble law for galaxy redshifts implied a uniform expansion - one in which every galaxy sees the same linear relation between distance and redshift when looking at other galaxies. The rate of this expansion is characterized by the Hubble constant - the ratio between redshift and distance for a fictitious average galaxy with no peculiar motion of its own with respect to the expansion. Not only does this value give us the size scale of the Universer, but it gives a measure of its age as well. If one runs the clock backwards on a uniform Hubble expansion, at a constant rate, the age of the expansion is the numerical inverse of the Hubble constant. Even if there has been deceleration of the expansion due to gravity, this age - the Hubble time - gives a scaling value for the age of the Universe.

The exact value of the Hubble constant has been contentious, with strong arguments presented for values from 50 to 100 km/sec per megaparsec. These correspond to Hubble times of, respectively, 20 and 10 billion years. Some of the disagreement between workers on the cosmic distance scale comes from different treatments of local galaxy motions superimposed on the smooth expansion, and some from regarding various measures of distance as primary or secondary. One of the major projects for the Hubble Space Telescope deals with direct measures of galaxies distant enough to expect a clean measurement of the Hubble constant.

The distribution of galaxies into groups, clusters, and superclusters carries information on masses and motions in the early Universe.  In brief, if the initial distribution of pre-galactic material was as uniform as the COBE satellite data suggests, then in order to form clusters today surrounded by relatively empty areas, the cluster-galaxies-to-be must have been able to move fast enough to cross (at least) the size of these empty areas in a Hubble time. The necessary gravitational clumping to propel such motion early enough proves to be an important constraint on the early Universe.

Classical cosmology was once described as a search for two numbers—the Hubble constant H0, and a second value, the deceleration parameter q0. The deceleration parameter described how fast the Hubble constant changes with cosmic time as the overall gravity of all matter in the Universe slows the expansion. A value of 0 would indicate an empty Universe—mathematically simple and appealing, but not very interesting to us! There are there different cases - an open Universe, in which the expansion will never stop; a closed Universe, in which the expansion will someday stop and reverse; and the critical point between them, where the expansion will constantly slow and approach (but never quite reach) zero. These are separated by the critical value q0 = . Many classical tests for the value of q0 relied on using galaxies as standard candles, but have been defeated by the fact that galaxies evolve on the same timescales that must be probed to measure q0. Current efforts in this direction use different probes, such as gravitational lensing or the mean mass density in the local Universe, in efforts to circumvent the unknowns of galaxy evolution. In any case, it is remarkable that q0 is close to its critical value; otherwise the Universe might not have the right properties for us to exist and discover cosmology.

[Gross simplification alert—specialists will complain, but we don't have enough room for a textbook here.]

Galaxy nomenclature and catalogs:

Galaxies are generally denoted only by catalog numbers; only a handful are well-known or unusual enough to rate distinctive names (such as the Whirlpool, Antennae, Pinwheel, and Cartwheel).  A given galaxy may sport numbers from several catalogs. The most cited sources are:

Messier number: from a list compiled visually bu Charles Messier and several colleagues during the eighteenth century. Many of the brightest and most conspicuous galaxies (as well as gaseous nebulae and star clusters) appear in the Messier lists.

NGC/IC (New General Catalog and Index Catalog): compilations by J.L.E. Dreyer from the 1860s-1880s. These included results of the complete sky sweeps performed by William and Jihn Herschel and discoveries by others, plus the first harvests of celestial photography. These catalogs include (besides the usual round of clusters and nebulae) about 10,000 of the most conspicuous galaxies. Until recently, almost all galaxies which could be studiesd in detail had NGC or IC numbers.

Arp: Halton Arp produced an atlas of peculiar and interacting galaxies, which first drew tha attention of many astronomers to the strange and spectacular forms that galaxies outside the normal Hubble classification can take.

UGC (Uppsala General Catalog): galaxies only, covering the sky north of –2 30'. Peter Nilsson produced this catalog of positions, sizes, orientations, and magnitudes from Palomar Sky Survey photographs.

Several other kinds of name comprise coordinate designations (first digits of the object right ascension and declination, either for epoch 1950 or 2000) and a survey name. Examples are the PKS (radio sources from the Parkes radio telescope in Australia) and IRAS (Infarred Astronomical Satellite) surveys. Thus we may have PKS 1413+003 or IRAS 09104+4109.

Many of the originally published versions of these catalogs are either rather obscure (university observatory transaction series, for example), or out of print (the NGC and IC being notable exceptions). Users of personal computers can get modern versions of these and many more on CD-ROMs (from the Astronomical Data Center at the NASA Goddard Space Flight Center, or the Almageste package from the ASP).  Such a level of access, without need of a professional astronomical  library, makes many kinds of advanced study and observing programs possible.

Note that wavelength/flux/surface brightness selection enters into what galaxies get selected for a particular catalog or study. Malin 1, despite being large and luminous, was long missed for being too large. Several researchers have pointed out that galaxy catalogs are dominated by those kinds of galaxies which are easiest to see against the natural glow of the night sky, so we may still be ignorant of important parts of the extragalactic census.

Observing galaxies

Even though most are difficult objects for visual inspection, the emotional impact of seeng "fossil" light makes galaxies popular targets for amateur astronomers. My own "best" list, including a few shown in the UA collection of WWW images, at, includes:

Name          Constellation   RA (2000)  Dec  Mag    Notes
-----------    ------------  -------- ------  --   --------------------------
M31=NGC 224   Andromeda      00 42.7  +41 16   4   Great spiral
M32=NGC 221   Andromeda      00 42.7  +40 52   9   Elliptical companion of M31
M110=NGC 205  Andromeda      00 40.4  +41 41   9   Elliptical companion of M31
M81=NGC 3031  Ursa Major     09 55.5  +69 04   8      
M82=NGC 3034  Ursa Major     09 55.9  +69 41   9   edge-on dusty starburst galaxy
Centaurus A   Centaurus      13 25.5  -43 01   8   peculiar radio galaxy
LMC           Doradus        05 23    -69 45   2   Large Magellanic Cloud
SMC           Tucana         00 52    -72 50   3   Small Magellanic Cloud
M77=NGC 1068  Cetus          02 42.6  -00 01  10   Seyfert nucleus
M87=NGC 4486  Virgo          12 30.8  +12 23   9   Center of Virgo cluster
NGC 4565      Coma           12 36.3  +25 59  10   edge-on spiral, dust lane
NGC 3556      Ursa Major     11 11.5  +55 40  11   edge-on spiral
NGC 891       Andromeda      02 22.5  +42 21  10   edge-on spiral
M104=NGC 4594 Virgo          12 40.0  -11 37   9   Sombrero galaxy
M65=NGC 3623  Leo            11 18.9  +13 05  10
M66=NGC 3627  Leo            11 20.2  +12 59  10
NGC 2903      Leo            09 32.1  +21 30  10   barred spiral
M83=NGC 5236  Hydra          13 37.0  -29 52   8   barred spiral
NGC 7331      Pegasus        22 37.1  +34 25  10   inclined, elongated
M94=NGC 4736  Canes Venatici 12 50.9  +41 07   9   very bright nucleus

For more of a challenge:
M33=NGC 598   Triangulum     01 33.8  +30 39   6   large but dim
Coma cluster  Coma Berenices 12 59.6  +27 57  13+
NGC 1300      Eridanus       03 19.7  -19 25  11   strongly barred spiral
NGC 147       Cassiopaeia    00 33.2  +48 30  11   distant companion of M31
NGC 185       Cassiopeia     00 39.0  +48 20  10   distant companion of M31
NGC 4038/9    Corvus         12 01.9  -18 52  10   "Antennae" interacting pair
NGC 6822      Sagittarius    19 44.9  -14 48   9   Local Group irregular

The whole region around M87, in the Virgo Cluster, contain numerous bright galaxies of various types. Take a good map.

As always, any particular size and magnitude listed for a galaxy may not gauge the visual impression for a given telescope, magnitude, and set of sky conditions. Look for yourself, your mileage may vary.

                       For Further Reading

The Hubble Atlas, by A. Sandage (Carnegie Institution of
   Washington 1961). This virtually defines the Hubble classification,
   and offers a stunning collection of galaxy photographs from the
   Mount Wilson and Palomar telescopes. It will soon be superceded
   by the Carnegie Atlas of Galaxies.
The Color Atlas of Galaxies, by J.D. Wray (Cambridge 1988). The fruits 
   of a long-term project to photograph galaxies through selected filters
   and produce composite color images. A wide range of galaxy types is
   illustrated, with care devoted to the color reproduction and its
Man Discovers the Galaxies, by R. Berendzen, R. Hart, and D. Seeley
   (Science History Publications 1976). From the recognition of
   galaxies through early distance measurements and the beginnings
   of modern observational cosmology. 
Galaxies, by Timothy Ferris (Sierra Club Books 1980). A classic 
   coffee-table book full of beautiful galaxy images. Useful even
   for career astronomers who need to remind themselves of what drew
   them into galaxy study.
Lonely Hearts of the Cosmos, by Denis Overbye. An engaging account of the
   quest for the cosmic distance scale.
The Universe of Galaxies, edited by Paul Hodge (Freeman 1984). Articles
    excerpted from Scientific American, covering dark matter, galactic
    tides, clusters, and active galactic nuclei.
Galaxies, by Paul Hodge

Source for the page:
(text originally produced for the Astronomical Society of the Pacific "Galaxies" slide set, by Bill Keel - see