FEATURE ARTICLE
The Formation of Star Clusters
Clouds in the summer sky provide clues about the organization of star populations
Bruce Elmegreen, Yuri Efremov
Globular Clusters, Old and Young?
The hierarchical structure of gas clouds and star clusters is apparent in the disk of the Milky Way, but there is another type of star grouping—the globular clusters found in the Galaxy's halo—that appears to challenge some of our ideas about the formation of clusters. Globular clusters are very compact and massive compared to open clusters. They may contain 100,000 to 1,000,000 stars within a volume of space no more than 20 or 30 light-years across. Remarkably, the globular clusters in the Milky Way's halo appear to be as old as the Galaxy itself, perhaps 10 to 15 billion years. For many years astronomers believed that the conditions needed to produce a globular cluster were no longer possible in today's universe. There are a few hints, however, that this is not the case.
In the Large Magellanic Cloud there are compact clusters that look very much like globulars. In 1951 Shapley and Virginia McKibben-Nail at Harvard University found a dozen Cepheid variable stars in one of the largest of these (NGC 1866). When this discovery was made it was surprising because it implied that the cluster is relatively young. The young age was also consistent with the blue color of the clusters. Hodge later confirmed that most of the Magellanic "globulars" were indeed quite young. These results were so disturbing to astronomers' ideas about globular clusters that they coined a new term for these objects, blue populous clusters.
Similar clusters have since been found in the disks of other galaxies, mostly in regions of intense star formation. Recently Bradley Whitmore of the Space Telescope Science Institute and Francois Schwiezer of the Carnegie Institution used the Hubble Space Telescope to uncover more than 1,000 luminous blue knots in the disks of two interacting galaxies, NGC 4038 and NGC 4039. These knots are as bright and blue as halo globular clusters would have been when they formed 10 to 15 billion years ago, but they are caught in the act of forming today, with the oldest age only a few hundred million years!
Some astronomers have argued that blue clusters in merging galaxies are unbound associations that will not evolve into old globular clusters. Like all of the disk clusters (bound or unbound) in our Galaxy, the young globular clusters in colliding galaxies have a scale-free distribution of masses. In contrast, old halo globular clusters have a very different mass distribution, one that is not scale-free but is centered around a characteristic mass of about 100,000 suns. This has prompted van den Bergh to suggest that the young luminous clusters in NGC 4038/4039 are more like Galactic-disk clusters than halo globular clusters.
This argument is difficult to refute, but another explanation for the centered mass distribution of old globular clusters is possible. Several astronomers, including Vladimir Surdin of the Sternberg Institute in 1979, Tadashi Okazaki and Makoto Tosa of Tohoku University in 1995, and ourselves, have conjectured that the whole distribution was born scale-free, but the low-mass clusters have since been destroyed, leaving today's distribution of masses peaked at some intermediate value. This scenario is consistent with theoretical work suggesting that low-mass globular clusters can evaporate or get destroyed by collisions with the galactic disk in a time shorter than the age of the universe, but the high-mass globulars can persist.
The high mass of the young globulars in interacting galaxies raises another question too, because all high-mass clusters in the disks of normal galaxies are unbound. Why should massive clusters be bound only in galactic halos and interacting galaxies?
» Post Comment