When trying to obtain an H—R diagram for a distant region of the Galaxy, astronomers face a problem. In order to plot the diagram, we must know luminosities; and in order to know luminosities, we must know distances. Thus it would seem impossible to construct H—R diagrams for stars more distant than 100 pc or so, the maximum distance measurable by stellar parallax. (We can't use the method of spectroscopic parallax because that method assumes the properties of the main sequence.)
In some circumstances, however, it is possible to plot an H—R diagram for very distant stars even though their distances are not known. If we observe a group of stars that all lie at the same distance from us, then comparing apparent brightnesses is equivalent to comparing absolute brightnesses. Why? Because as the radiation travels toward Earth, the brightness of every star in the group is diminished by the same amount, according to the inverse-square law. By measuring and plotting apparent brightnesses, we can create a "relative" H—R diagram for the group that looks (apart from the numbers on the vertical axis) exactly the same as a real H—R diagram based on actual luminosities. One such easily recognizable group of distant stars is called a star cluster.
Star clusters can include anywhere from a few dozen to a million stars in a region a few parsecs across. Astronomers believe that all the stars in a given cluster formed at the same time, out of the same cloud of interstellar gas, and under the same conditions. Thus, when we look at a star cluster, we are looking at a group of stars that all have the same age, are similar in composition, and lie in the same region of space, at essentially the same distance from Earth. Unlike the stars plotted in Figures 17.13—17.15, which differ in mass, age, and (to a lesser extent) chemical composition, the only factor distinguishing one cluster star from another is its mass.
Clusters are therefore almost ideal "laboratories" for stellar studies—not in the sense that astronomers can perform experiments on the stars in them, but because the properties of the stars are very tightly constrained. Hence theoretical models of star formation and evolution can be compared with reality without the major complications introduced by broad spreads in age, composition, and place of origin. Clusters are of central importance to astronomers who wish to understand how stars evolve in time.
Figure 17.23(a) shows a rather loose cluster—the Pleiades, or Seven Sisters—a well-known naked-eye object in the constellation Taurus. Individual stellar colors provide an estimate of the surface temperature of each star in that cluster. Luminosities follow directly from measurement of the apparent brightness and the cluster's distance (which in this case is known to be about 120 pc). Figure 17.23(b) shows the cluster H—R diagram obtained from these data. This type of cluster, found mainly in the strip across the sky known as the Milky Way, is called an open cluster. Open clusters typically contain from a few tens to a few hundred stars and are a few parsecs across.
Figure 17.23 (a) The Pleiades cluster (also known as the Seven Sisters, or M45) lies about 120 pc from the Sun. The naked eye can see only its brightest stars. (b) The stars of this well-known open cluster yield an H—R diagram.
The H—R diagram in Figure 17.23(b) shows stars throughout the main sequence—stars of all colors are represented. The blue stars must be relatively young, for, as we have seen, they burn their fuel rapidly. If all the stars in the cluster formed at the same time, then the red stars must be young too. Thus, even though we have no direct evidence of the cluster's birth, we can estimate its age as less than 25 million years, the lifetime of an O star. Other factors also hint at the cluster's youth. It contains a large amount of interstellar gas and dust not yet processed into stars or lost from the cluster. And it is abundant in heavy elements that (as we will see) could have been created only within the cores of many generations of ancient stars long since perished.
A second type of stellar cluster, of which a representative is shown in Figure 17.24(a), is called a globular cluster. Globular clusters are much more tightly knit than the loose groups of stars that make up open clusters. All globular clusters are roughly spherical (which accounts for their name) and contain hundreds of thousands, and sometimes millions, of stars spread out over about 50 pc. As with open clusters, the entire assemblage is held together by gravity.
Figure 17.24 (a) The globular cluster Omega Centauri is approximately 5000 pc from Earth and spans some 40 pc in diameter. (b) An H—R diagram for many (but not all) of its stars.
Figure 17.24(b) shows an H—R diagram for this cluster, which is called Omega Centauri. Notice its many differences from Figure 17.23(b)—globular clusters are a very different stellar environment from open clusters like the Pleiades. The distance to this cluster has been determined by a variation on the method of spectroscopic parallax but applied to the entire cluster rather than to individual stars. From calculations of the distance at which the apparent brightnesses of the cluster's stars taken as a whole best match theoretical models, the cluster is found to lie about 5000 pc from Earth.
The most outstanding feature of globular clusters is their lack of O- and B-type stars. Astronomers in the 1920s and 1930s, working with instruments incapable of detecting stars fainter than about 1 solar luminosity at such distances, and having no theory of stellar evolution to guide them, were very puzzled by the H—R diagrams they saw when they looked at globular clusters. Indeed, a comparison of just the top halves of the diagrams (so that the lower main sequences cannot be seen) reveals few similarities between Figures 17.23(b) and 17.24(b).
We now know that, although low-mass red stars and intermediate-mass yellow stars abound, globular clusters contain no main-sequence stars with masses greater than about 0.8 the mass of the Sun. (The A stars in this plot are stars at a much later stage in their evolution that happen to be passing through the location of the upper main sequence.) Apparently, globular clusters formed long ago; the more massive O through F stars have long since exhausted their nuclear fuel and disappeared from the main sequence (becoming the red giants and other luminous stars above the main sequence, as we will discuss in Chapter 20. Other factors confirm that globular clusters are old. For example, their spectra show few heavy elements, implying that these stars formed in the distant past when heavy elements were much less abundant than they are today.
On the basis of these and other observations, astronomers estimate that all globular clusters are at least 10 billion years old. They contain the oldest known stars in the Milky Way Galaxy. As such, globular clusters are considered to be remnants of the earliest stages of our Galaxy's existence.
We will never be able to watch a single star move through all its evolutionary phases. The lifetimes of humans—even of human civilizations—are far too short compared with the lifetimes of even the shortest-lived O and B stars. Instead, we must observe stars as they presently exist—through "snapshots" taken at specific moments in their life cycles. The H—R diagram is just such a snapshot. By studying stars of different ages or, even better, by studying stars in a cluster, in which the ages are known to be the same, we can patch together an understanding of a star's "life story" without having to follow a few individuals from birth to death. Such evolutionary studies will be the subject of the next few chapters.
