Modern astronomy paints a vivid picture of the universe having been born in a cataclysmic bang and filled with exotic stars ranging from gargantuan red supergiants that span the size of a modest solar system to hyperdense white dwarf stars and black holes that are smaller than Earth. These discoveries are all the more remarkable because astronomers infer them from the faintest glimmers of light, sometimes just a handful of photons. A key to this success is a graph that two astronomers introduced 100 years ago.

The Hertzsprung-Russell (H-R) diagram is simple. It plots two basic properties of stars: their luminosity (intrinsic brightness) and their surface temperature (as revealed by their color). In doing so, it anchors stellar astronomy just as the periodic table anchors chemistry. Whereas the periodic table groups together similar chemical elements—for example, placing all noble gases, such as helium, neon and argon, into one column—the H-R diagram groups together stars passing through similar stages of life. When astronomers invented the diagram, no one knew why the sun and other stars shine. No one knew how stars are born or how they die. No one could even assure the public that the sun would never explode. Nor did anyone know that the stars had forged most of the elements that make up Earth and our bodies.

Not only did the diagram play a major role in solving these problems, it also guides astronomers today as they tackle key questions about the stars. How massive can a star be? What were the first stars to arise after the big bang? When will we see the next supernova in our galaxy?

A Tour of the Stellar Bestiary
“Nobody imagined that I should become an astronomer,” said Danish scientist Ejnar Hertzsprung. Indeed, when he was 20, his family sold his late father’s astronomy books. Nevertheless, Hertzsprung persevered. He sketched his first luminosity-color diagram of star clusters in 1908. German astronomer Hans Ros­en­berg, who likely knew of Hertzsprung’s work, published such a diagram in 1910, and Hertzsprung himself published several in 1911. At the time, he was an unknown. In contrast, Henry Norris Russell was one of America’s foremost astronomers. In 1913, unaware of Hertzsprung’s work, he plotted his own diagram. Because of Russell’s prestige, astronomers first called the plot the Russell diagram, then the Russell-Hertzsprung diagram and finally—getting the historical order right—the Hertzsprung-Rus­sell diagram.

As astronomers plotted stars on the graph, they discovered clear patterns. The vast majority, including the sun, lie on a diagonal line stretching from the upper left (bright and hot stars) to the lower right (dim and cool ones). This diagonal, which astronomers call the main sequence, is a startling revelation, because it links stars that seem to be opposites. Every main-sequence star generates light the same way: nuclear reactions convert hydrogen into helium at the star’s center. The more mass a main-sequence star has, the hotter its center gets and the faster the reactions proceed, making the star brighter and hotter. Thus, the main sequence is really a mass sequence.

Another stellar group appears above and to the right of the main sequence. It consists of stars that are brighter than main-sequence stars of the same temperature and color. Most are cooler than the sun; all are brighter. At first that sounds like a contradiction: the cooler a star is, the less light every square inch of its surface radiates, so how can a cool red star shine 100 or even 10,000 times more brightly than the sun? The answer is that these stars must be enormous—astronomers call them giants and supergiants. They are what main-sequence stars become after they exhaust the hydrogen fuel at their centers. Supergiants eventually explode as supernovae. Giants exit the scene more quietly.

In fact, the H-R diagram reveals the fate of the giants. The diagram contains a group of stars that form a diagonal line below the main sequence, which means they are dimmer than main-sequence stars of the same temperature and color. By the same reasoning as discussed, these stars must be tiny—astronomers call them white dwarfs. Despite their name, they stretch across many colors. They are the dense and intensely hot cores left behind when giants cast off their outer atmospheres. No longer capable of nuclear reactions, they usually cool and fade with time. If they are part of a binary star system, however, they can suck in matter from their companion star, reach a critical mass and go supernova.

The distinctive and ubiquitous patterns of the H-R diagram even reveal stellar properties that the diagram does not directly display. For example, astronomers can ascertain the age of a star cluster by plotting an H-R diagram just for the stars in that cluster. In the Pleiades cluster, for example, the main sequence extends to bright blue stars, whereas in the Hyades, such stars are missing. Consequently, the Hyades must be older; the bright blue stars it used to contain have all died off.

Bigger and Badder
The H-R diagram remains a vital tool. Much of today’s research in stellar astronomy can be thought of as a way to explore the extremes of the diagram. At the bottom right are the dimmest, reddest, least massive stars. The main sequence ends with dim red stars that have about 8 percent of the sun’s mass. Beyond is the realm of brown dwarfs, stars that are too lightweight to sustain nuclear fusion. Their properties and genesis still puzzle astronomers [see “The Mystery of Brown Dwarf Origins,” by Subhanjoy Mohanty and Ray Jayawardhana; Scientific American, January 2006].

At the other end, the upper left of the H-R diagram is the home of the brightest, hottest, most massive main-sequence stars. But how massive can they get? Bright stars are easy to see but difficult to study because they are rare. Few are born, and those few burn their fuel so fast that they explode a few million years after birth. Studies of very young star clusters suggest that stars top out at about 150 times the sun’s mass. Last year, however, Paul Crowther of the University of Sheffield in England and his colleagues upped the ante. They claimed that a star in the Large Magellanic Cloud, a modest near­by galaxy, was so bright and blue that it must have been born with a whopping 320 solar masses. Some astronomers are skeptical about the mass estimate, however, because it assumes that the star follows the same pattern of mass, brightness and temperature as ordinary main-sequence stars.

Whatever the case, the very first stars in the universe may have been even larger. The big bang created the three lightest elements: hydrogen, helium and a little lithium. The primordial soup lacked carbon and oxygen, which emit infrared light that escapes present-day interstellar clouds and thereby allows them to cool and fragment. Thus, the first star-forming gas clouds may have been warm and large, and they should have given birth to stars with hundreds of times the mass of the sun [see “The First Stars in the Universe,” by Richard B. Larson and Volker Bromm; Scientific American, December 2001]. If so, they were much brighter and hotter than the most extreme stars today; they would therefore appear above and to the left of the upper left corner of the modern H-R diagram.

Any star born with more than eight times the mass of the sun someday explodes [see “How to Blow Up a Star,” by Wolfgang Hillebrandt, Hans-Thomas Janka and Ewald Müller; Scientific American, October 2006]. Every year astronomers witness hundreds of supernova explosions in galaxies beyond our own. But not since 1604—before astronomers were using the telescope—have they witnessed a star go supernova in our galaxy. Which will be the next to self-destruct, and when will we see it?

The Milky Way spawns a couple of supernovae a century. But when one goes off, there is no guarantee we will see it. The Milky Way is vast—far larger than most other galaxies—and its disk is choked with interstellar dust, which blocks the light even of a supernova. Indeed, more than half a century ago astronomers discovered a giant cloud of debris named Cassiopeia A; the light from the explosion that created it reached Earth in the late 1600s but went unnoticed.

Thus, any exploding massive star that makes a splash in the sky will have to be nearby, probably within about 20,000 light-years of Earth. To find stars on the brink, astronomers look in the upper right of the H-R diagram—the realm of the red supergiants. The nearest and brightest are Betelgeuse and Antares, which are 640 and 550 light-years from Earth, respectively—close enough that their explosions will rival the moon in brightness but far enough that they should not hurt us.

But the cosmos can always surprise us. The famous 1987 supernova in the Large Magellanic Cloud came not from a red supergiant but from a blue one. Similar stars also reside in our galaxy; they include two of the most conspicuous stars in the night sky, Deneb and Rigel.

Or we could see another type of supernova, which results when a white dwarf exceeds a critical mass. Although such supernovae are rarer, they are also more luminous and usually occur above or below the dusty disk, making them easier to see. Of the five supernovae in our galaxy astronomers have seen since a.d. 1000, three—and possibly four—were exploding white dwarfs. Unfortunately, white dwarfs are so dim that the suspects for triggering the next supernovae are not obvious.

Nevertheless, light from the next Milky Way supernova is racing toward us right now. When it finally arrives, astronomers will plot the progenitor’s position on the H-R diagram to understand its life and death. Hertzsprung and Russell would be pleased to know that their creation still yields so much insight. Moreover, its success has inspired similar plots of other phenomena, notably, the many planets orbiting other stars. Such a graph may unveil as much about Earth’s galactic relations as the H-R diagram has revealed about the sun’s.

A User’s Guide to the Periodic Table of the Cosmos

Stellar Color and type
The color of a star reflects the temperature of its surface, from tepid red-hot (far right) to sizzling blue-hot (far left). Astronomers divide stars into seven main spectral types, based on which chemical elements in their outer layers absorb light, which in turn depends on temperature: O, B, A, F, G, K and M. The universal mnemonic is “Oh, be a fine girl/guy, kiss me!” although the alternative “Oh, boy, an F grade kills me!” has its appeal.

Main sequence
Most stars fall in a diagonal line, indicating that their luminosity and temperature are determined by a third, even more basic property: mass. The hot, bright stars on the left are the most massive. Once a star begins producing energy by fusing hydrogen nuclei, it achieves a stable internal equilibrium and stays near the same spot on the diagram for most of its life. 

These are ex-main-sequence stars that have exhausted the hydrogen in their core and now devour other reservoirs of fuel, such as helium. The most massive become supergiants; lesser ones, giants. If a large red supergiant replaced the sun, it would engulf all the planets out to Jupiter. These stars do not remain at a fixed position on the diagram but move around as they age. 

The most massive stars of all are found near the top of the diagram. The current record holder is R136a1, which, at birth, was 320 times as massive as the sun; since then, it has lost  mass by expelling gas. A similarly massive and unstable star is Eta Carinae, which is enveloped in a gaseous nebula from an outburst 170 years ago.

White dwarfs
White dwarfs are stellar corpses. Unable to generate energy anymore, they pack themselves into balls barely the size of Earth. Their name notwith­standing, they span a range of colors. Over time a white dwarf slips down the chart to the right, until it can barely be seen. The star is initially swaddled in a so-called planetary nebula consisting of ejected gas.

The sun lies on the main sequence. It came into being as a cool protostar and, once it exhausts its core’s hydrogen fuel, will become a red giant and finally a white dwarf. Contrary to popular belief, the sun is not an average star; some 95 percent of stars lie below it in the diagram. This ultra­violet image shows a solar prominence in September 1999.

Brown dwarfs
A frontier of astronomy in the past two decades has been the detection and study of brown dwarfs, which are stars too light to undergo sustained nuclear fusion. On the diagram, they overlap with the dimmest, reddest stars at the bottom right and continue off the page to the right. LP 944–20 in the bottom right corner is one. A decade ago astronomers added spectral types L and T (not shown) to categorize these objects.