INTRODUCTION

A star cluster is group of stars that formed out of the same molecular cloud, and which remains bound together by the mutual gravitational attraction of its components. Clusters have been traditionally classified as either open or globular, based primarily on their visual appearance (and additionally confirmed by their underlying spectral signature).

How do each of these types of clusters evolve? What are their respective populations, and their dynamics? What do we know about them? And most of all: Do we lack anything in our knowledge of star clusters? What about future research?

The Pleiades open cluster contains some six or seven components which are easily seen with the naked eye, forming an irregular pattern somewhat resembling the shape of a dipper. The total count of stars actually reaches about 500 (Freedman & Kaufmann, 2002), spread throughout an area with a diameter of 110 arc minutes. The ESA Hipparcos mission has made an accurate determination of its distance (379 10 light years), making this cluster one of the few with a precise determination of this crucial parameter.

Spectral analysis of the Pleiades cluster reveals a whole sample of main-sequence stars, ranging from hot, luminous B stars, to cold and faint M stars, as well as a relative abundance of metals (that is, elements heavier the hydrogen or helium) present throughout virtually every star. The cluster is located near the galactic plane, orbiting the center of the Galaxy along the disk.

Globular cluster Omega Centauri presents a quite different scenario. Visible to the naked eye as a small nebulous patch of about 4^th magnitude, even a small telescope is able to reveal its real glory: a vast, spherical array of about a million stars (Kaler, 1998) concentrated over an area of only 36 arc minutes in diameter. Its distance has been estimated at approximately 16,000 light years (SEDS).

Spectral measurements of selected components of Omega Centauri show a peculiar distribution along the Hertzprung-Russell diagram: hot, blue-white stars seem rather scarce, with a large number of cluster components laying along the yellow-orange-red part of the spectral sequence. Its metallicity index of 1.59 (Hirshfeld & Sinnott, 1985) in turn reveals a significant scarcity of metals. This object is located at a considerable distance above the galactic plane, orbiting the bulge of the Galaxy from within the spherical outer halo.

What does this tell us about clusters in general?

The Hertzprung-Russell diagram of an open cluster will normally reveal a wide variety of main sequence stars from over the whole spectral range, except for old clusters, which have mostly lost their massive O and B stars, as well as their low-mass K and M stars [see figure 1].

Figure 1. Hertzprung-Russell Diagram of a Young Open Cluster Copyright 2003 Armando Caussade. Comments: Spectral types are marked along the X-axis, while absolute magnitudes are marked along the Y-axis. These values express a star's surface temperature and luminosity, respectively.

Open clusters originate in molecular clouds located within the disks of spiral and barred-spiral galaxies. As these clouds orbit around the galactic center, they will inevitably encounter spiral arms or bars, where they will be compressed, thus triggering the birth of new clusters of stars (Freedman & Kaufmann, 2002). As these newly-born clusters mature, they will continue their orbits around the galactic center, encountering on their way a number of field stars, or even a molecular cloud. These perturbations will tend to eventually disassociate open clusters until they finally scatter along the galactic disk. These forces are felt with particular strength by low-mass stars, which is usually why the K and M stars are the first to be lost.

Over 1,100 open clusters have been identified in our galaxy, with a probable total number as high as 100,000 (SEDS). Well known examples of open clusters located inside our own Milky Way Galaxy include the Trapezium Cluster (associated with the M42 nebula, in the constellation Orion) and M67 (in the constellation Cancer). The former is an extremely young cluster, populated mainly by hot blue stars, where star-forming activity continues even now, while the latter is a very old cluster born about about 3,200 million years ago, and located far above the galactic plane. (Hirshfeld & Sinnott, 1985).

Many extragalactic open clusters (that is, clusters associated to other galaxies) have been found, the most notable being the 30 Dorads cluster (NGC 1910). This object is considered the most luminous (and massive) young cluster in our Local Group of Galaxies (Larson, 1993, 1996). 30 Dorads is associated with the gigantic Tarantula Nebula (NGC 2070) a gigantic star-forming complex containing an estimated 100 million solar masses of gas and dust.

Then, how does this compare to globular clusters?

The Hertzprung-Russell diagram of a globular has a shape where the upper half of the main sequence is practically absent. This is where, in fact, the hot, early-type stars (O, B, and A stars) concentrate. In its place, there is a prominent horizontal branch centered at around magnitude 0 [see figure 2].

Figure 2. Hertzprung-Russell Diagram of a Globular Cluster Copyright 2003 Armando Caussade. Comments: Spectral types are marked along the X-axis, while absolute magnitudes are marked along the Y-axis. These values express a star's surface temperature and luminosity, respectively.

Globular clusters are almost always found outside of the galactic disk, orbiting around the Galaxy from within the halo at high inclinations. While their stars will frequently remain bound together for long periods of time, they will inevitably suffer from a variety of orbital perturbations, particularly, tidal forces originating in the galactic bulge. This will cause a slow attrition pattern that will lead to their eventual disruption or "evaporation". It is estimated that half of our Milky Way's globulars will disappear during the next 10,000 million years (SEDS).

The origin of globular clusters is the one big question which still remains unanswered. It has been suggested that a number of "young compact star clusters" detected around other galaxies could be described as "protoglobulars" (Whitmore, 2000), but this needs to be further researched. Whatever these objects turn out to be, it seems apparent that new globulars are now an exceedingly rare occurrence, and that the time for globular cluster formation has already passed.

Yet the phenomenon of globular accretion by galaxies seems to be rather common. It is thought that four "new" globular clusters (previously belonging to the Sagittarius Dwarf Galaxy), including the massive M54 (Larson, 1996) may actually be in the process of being incorporated within our own galactic halo, after an extremely close passage by the Sagittarius Galaxy.

Around 150 open clusters have been identified in our galaxy, with a probable total number of about 180 to 200 (SEDS). Examples of globular clusters located around our own Milky Way Galaxy include 47 Tucanae (in the constellation Tucana) and M13 (in the constellation Hercules). The Tucana cluster is plainly visible to the naked eye; it is highly concentrated, and has been found to have a relatively high metallicity index. M13, on the other side, has been found to have an overall or composite spectrum which is earlier than both 47 Tucanae and Omega Centauri.

Many globular clusters have been seen around other galaxies, with sometimes thousands orbiting a single galactic system. A high frequency of globulars has been observed particularly around elliptical galaxies (Larson, 1996). It is thought that some galaxies of this kind may have actually formed by merging with other systems, which would help explain the abundance of globular clusters around them.

But, are open and globular clusters really so different?

Yet, even considering these significant differences, some researchers have recently suggested that both types of clusters could be better described as a single continuum of objects, rather than separate, discrete categories (Larson, 1993, 1996). Physical distinctions between these two types of clusters (whether inferred or observed) would thus be quantitative, rather than qualitative. Perhaps, a better knowledge on the birth and early evolution of globular clusters will help to answer this intriguing question.