"In the case of a single star, its brightness and temperature don't tell us much. Because these properties stay fairly constant for 90 percent of its lifetime, the star could be fairly young or fairly old, and we wouldn't be able to tell the difference. In a star cluster, we have the advantage that stars of all masses formed at about the same time. So all we have to do is look at the cluster and determine how hot and how massive is the hottest, bluest, most massive star that has not yet entered the late, unstable period of its life. The star's mass tells us how much fuel the star had when it was born, and the star's brightness tells us how fast it is burning that fuel. We know that the star is just about to start becoming unstable--after all, the stars that are more massive have already started to become unstable. We also know that its fuel is just about exhausted. The ratio of how much fuel the star had in the beginning to how fast it has been burning that fuel tells us how long the star has been alive. (By analogy, if we know how much kerosene our hurricane lamp contained when we lit it and how fast it consumes the kerosene, and if the lamp is just now starting to go out, then we can deduce how long it has been lit.) Because all the stars in the cluster are the same age, the age of that one star tells us the age of the entire cluster.
"The basic physics of how hydrogen is converted to helium in the centers of stars and the amount of energy generated by this process is comparatively simple and well understood. For much of the 20th century, the main limitation to our knowledge of stellar ages has been due to the difficulty of measuring the distances to the clusters--especially the distances to the oldest clusters, the globulars, which are comparatively far away. (We know how bright a star looks, but to know how bright it really is, you have to know how far away it is: is it like a headlight a mile away or an airport beacon 10 miles away? In the dark of the nighttime sky with no reference points, it's pretty hard to tell.) Technical advances, such as the introduction of charge-coupled devices to replace photographic plates for the measuring of stellar distances and brightnesses, are making our observations more secure.
"Distance measurements have improved to the point at which other details needed to determine the ages of star clusters--such as the fine details of how a star converts nuclear energy to visible light--can no longer be ignored. How exactly does the energy get from the center of the star, where it is generated, to the surface, where it becomes the light that we see? How important is convection as a means of transporting energy, and how efficient is the convection? The answer to these questions has some effect on the inferred relationship between mass and surface temperature. Just how much oxygen is in the stars, along with the hydrogen and helium? The relative amount of oxygen present has a modest effect on the efficiency of the central furnace, affecting the relation between mass and brightness and, hence, age.
"Taken together, the uncertainty in the observations and the uncertainty in the relevant theoretical physics probably lead to an uncertainty of 10 percent to 20 percent in our estimate of the absolute ages of the globular clusters. According to our best available estimates, stars having about 90 percent of the sun's mass are just now starting to die in the globulars. These stars are most probably around 15 billion years old, but they could conceivably be as young as 12 billion years or as old as 18 billion years. It is very unlikely that most of them could be either younger or older than this range. This estimate is already accurate enough to place some very interesting limits on the age and life history of the universe."