I remember, as a child, going to the Exploratorium in San Francisco. The Exploratorium is a hands-on science museum for kids and adults, where you can investigate such disparate science topics as the motion of objects in a gravitational field, the effect of darkness and delay on three-dimensional vision, how engines work, and so forth. It's still there, and a great way to spend a day.
One of the exhibits they had was a sort of primitive pinball machine, with only one bumper. This bumper had an unknown shape -- at least, it was unknown to the user at first -- because it was hidden underneath a circular mask, which covered it completely. However, you could fire pinballs up at the bumper, which would strike it underneath the mask, and then bounce back every which way. You could slide the plunger back and forth along the bottom edge of the frame, and also rotate the bumper, so that you could strike the bumper from different positions and angles.
The purpose of all this firing and bouncing was to try to determine the shape of the bumper. You couldn't see the actual bumper, you couldn't get at it to take it apart; all you could do was fire pinballs up at it and observe which way they bounced. They had a whole bunch of these machines; eventually, you could piece together that the shape was a star, or a hexagon, or a triangle, or whatever it was. Sometimes you would decide right away that it was a square, and all of a sudden one of the pinballs, fired just right, would bounce back in a completely unexpected direction. You would have to play around a little longer before you realized that the shape was really a cross.
That, to me, is the essence of science. You practically never have full and complete access to the internal workings of whatever it is you're trying to figure out. Either it's too small, or too old and decayed, or too hard to get at; it's always too something to look at directly and figure out what's going on. You're always limited in how you can manipulate the darn thing. And yet, even using a limited tool set, you can arrive at a more complete picture of how it works.
In astronomy, of course, the problem is that everything is too far away. Only in this century has it become feasible to travel to the moon and planets, and even then we've only been able to bring back samples from the moon for first-hand observation and analysis. And although it now appears conceivable to one day travel to the stars, it's hardly out of the question that we still won't be able to go in and take core samples. How, then, do we figure out what's in the stars, and what makes them shine?
For generations upon generations, human observers have looked up at the stars and been unable to see anything more distinct than dots of light. It might have been possible for eagle-eyed humans to see the discs of Venus or Jupiter, to be sure, but even the ancient Greeks knew that these were planets and not fixed stars, so of course they were different. The stars remained infinitesimal points of light.
Even after Galileo first brought a telescope to bear on the stars (and no, he did not invent it), that's all anyone ever saw: points of light. Galileo's telescopes were so inferior by today's standards that he was unable to distinguish Saturn's rings, something that practically any dime-store telescope nowadays can do. Anyone who saw the stars as finite in extent soon discovered that their telescope was in error, and that they had not actually seen the surface of a star. Centuries passed, telescopes improved in both resolution and light-gathering power -- and points of light the stars steadfastly remained.
In the early 19th century, the French philosopher Auguste Comte (1798-1857) was trying to come up with an example of something that would forever be unattainable. The inability to resolve even one star other than our own sun led him to conclude that we would never be able to determine the composition of the stars. We would never be able to sample them physically, he argued, and there seemed no other way to analyze them, since no one could see the surface of any star, other than our own sun:
To attain a true idea of the nature and composition of [astronomy], it is indispensable. . .to mark the boundaries of the positive knowledge that we are able to gain of the stars. . . .We can never by any means investigate their chemical composition. . . .The positive knowledge we can have of the stars is limited solely to their geometrical and mechanical properties. [Cours de philosophie positive, 1842]
By "mechanical properties," Comte meant the movement of the stars through space, not their internal workings. As it sometimes goes, though, it took only a couple of years after Comte died for him to be proven wrong, and it didn't in fact require travel to the stars. Since Isaac Newton (1642-1727), it had been known that white light, such as light from the sun, could be separated into its constituent colors using a prism. This spectrum, as it came to be called, appeared to be continuous -- that is, there were no gaps anywhere where colors might be missing.
By the 19th century, however, the Bavarian optician Joseph von Fraunhofer (1787-1826) had discovered that the spectrum was not, in fact, completely continuous, but instead had little gaps here and there. These gaps are now called Fraunhofer lines in his honor. Then, in 1859, the German physicist Gustav Kirchhoff (1824-1887) discovered that when you heated certain minerals, the light emitted by the resulting flame wasn't a smooth and complete spectrum, as was light from the sun. Instead, it was a collection of small bands of light, each in its proper position along the spectrum, but in isolation.
Could it be that the gaps and the light bands were related in some way? When Kirchhoff looked at a pair of gaps in the solar spectrum, near the orange portion, he noticed that they seemed to be in exactly the same position as the two lines emitted by burning sodium. To test their similarity, he decided to pass sunlight through burning sodium vapor, and then subject the combined light to a prism, expecting to see the emission lines of the sodium vapor fill in the gaps in the solar spectrum. To Kirchhoff's great surprise, they did no such thing. If anything, the gaps became even darker and more distinct than they had been without the sodium vapor. Were the emission lines of the sodium vapor just off a bit from the gaps in the solar spectrum, and somehow mysteriously drawing further light from those gaps?
It took a few experiments more before Kirchhoff discovered what was going on. The burning sodium vapor was in fact emitting the same lines as were missing in the solar spectrum. However, when light from one source (the sun) passes through another light source cooler than the first (the burning sodium), the second source absorbs precisely the same lines it would emit by itself. In recognition of Kirchhoff's investigations into gas emissions and absorptions, this whole business about cooler gases absorbing and hotter gases emitting is known as Kirchhoff's Law. In this way, he had proven Comte wrong, just two years after the latter's death.
Soon after, in 1862, the Swedish physicist Anders Ångstrom (1814-1874) examined the solar spectrum in much finer detail than Kirchhoff had, and discovered hydrogen in the sun. Ångstrom measured the position of the hydrogen lines (and a whole host of others) to a precision of one ten-billionth of a meter. In his honor, that unit of length is now named after him, and you will still see, occasionally, light wavelengths measured in Ångstroms. (The appropriate official metric unit is the nanometer, which is a billionth of a meter.)
He had to measure it so finely, because, as it turns out, hydrogen produces relatively faint absorption lines, even though it constitutes most of the matter in the sun (and, indeed, the universe). The sun, we now know, is about 75 percent hydrogen. The remainder is mostly an element that was first discovered six years later when the French astronomer Pierre Janssen (1824-1907) detected absorption lines that didn't appear to correspond to any known element. The British astronomer Joseph Lockyer (1836-1920) examined Janssen's data, agreed that the element indicated was a previously unknown one, and named the newly discovered substance helium, after the Greek word for "sun."
Hydrogen and helium are the two lightest elements, which explains why the earth has so little of them. The earth's gravity is simply insufficiently strong to hold onto hydrogen and helium -- it can only hold onto heavier gases, such as nitrogen, oxygen, carbon dioxide, and so forth. Hydrogen is at least reasonably reactive, so that it forms lots of compounds that the earth can hold onto, such as methane and ammonia and water. Helium, on the other hand, is an inert gas; it tends to "ignore" other elements, and the only reason that we have any amount of it for blowing up zeppelins and party balloons is that it is a byproduct of the decay of radioactive elements. (There's an interesting story about how helium was first discovered on the earth, but that's a question for another time.) When a party balloon loses its helium, that helium dissipates into the atmosphere, where it eventually finds its way outward into space and is lost forever. The sun, on the other hand, is so massive that although it is very hot, so that the hydrogen and helium atoms are jostling about very rapidly, its gravity is more than sufficient to keep those atoms from escaping.
So, the sun is made of about three-quarters hydrogen and one-quarter helium, and a tiny smattering of other elements. If we mix those gases here on earth, we certainly don't get the sun. We could light it up, and then the mixture would likely go up in a blaze of fire (provided enough oxygen were present), but it would last only for a brief time. How then does the sun stay lit?
People have been wondering about that for a long time. The earthly activity most like the shining of the sun is clearly the burning of fire. Here, of course, fire is dynamic, local in effect, and temporary in duration, whereas the sun seems steady and wide-ranging, and as far as anyone can tell, it's been shining essentially forever. Nevertheless, this was a reasonable analogy on the face of it, so before we knew what the sun was made of, the German physician Julius Mayer (1814-1878) calculated that if the sun were essentially a humongous lump of coal, there was enough fuel there to last about 5,000 years. Other substances could be substituted in place of coal, but none of them, in fact no chemical phenomenon at all, could be relied on to produce the sun's level of energy production for more than several thousand years, and in any case the soon-to-be-discovered composition of the sun made most of these infeasible to begin with.
Mayer himself then proposed another explanation. Perhaps, he suggested, meteoroids and other space debris were attracted continually by the sun's gravity and fell into the sun, in the process generating heat. It had been known for some time that a moving object carries kinetic energy according to the equation:
KE = (mv^2)/2
where m is the object's mass and v is its velocity.
When an object comes to a stop, that kinetic energy cannot simply vanish. It has to be transferred elsewhere, and the "lowest common denominator" form of energy is heat. This explains why a hammer gets hot after striking a lot of nails -- because the hammer's motion is impeded by the nails (or, if you're unlucky, your thumb), and the kinetic energy of the hammer's bulk motion is transferred to the kinetic energy of the molecules in the hammer (and the nails and your thumb, too, for that matter). Heat, in other words. Something similar might be happening with the sun, and if so, maybe that was keeping the sun lit.
Mayer's proposal had the advantage of explaining how the sun could continue shining indefinitely, since as far as anyone knew, there could be a limitless supply of space debris. However, the Irish physicist William Thomson (1824-1907), later Lord Kelvin, discovered another problem with Mayer's proposal. When space debris collides with the sun, its kinetic energy is turned into heat and that can be radiated away. However, the amount of kinetic energy, according to the above formula, depends not only on the velocity of the debris but on its mass as well, and that mass cannot be radiated away. It has to stay in the sun's bowels. Thomson calculated the rate of mass that would have fall into the sun to support its current power output, and determined that the sun would have to "gain weight" so quickly that its gravitational force would have increased and the orbit of the earth would have shrunk measurably in historic times. Since no such shrinking had ever been detected, Mayer's proposal had to be rejected.
Well, if falling objects couldn't keep the sun lit, perhaps the sun was falling in on itself. The German philospher Immanuel Kant (1724-1804) had earlier proposed the nebular hypothesis of the sun's formation, which theorized that the sun and the rest of the solar system condensed out of a great cloud of gas and dust. If Kant was right and the sun had indeed collapsed from a cloud of gas and dust, maybe that condensation itself was sufficient to power the sun. That way, the sun could heat up without gaining mass and changing the earth's orbit in any way. The German physicist Hermann von Helmholtz (1821-1894) calculated that under reasonable assumptions, this process was enough to keep the sun radiating at its current power for several million years. Unfortunately, by this time, the earth was known (from geological and biological lines of reasoning) to be much older than this.
The crucial clue came in the form of the famous theory of general relativity, developed by the Swiss physicist Albert Einstein (1879-1955). In it, Einstein developed his famous equation,
E = mc^2
which tells us that any mass -- a bit of starstuff, an automobile, dryer lint -- can be transformed into an astounding amount of energy. The British astronomer Arthur Eddington (1882-1944) suggested that the sun had so much hydrogen and helium because it was turning the hydrogen into helium. Helium was available in small quantities to study in the laboratory, and it had been discovered that a helium atom weighed almost but not quite as much as four hydrogen atoms; the precise ratio was closer to 3.97 to 1. Perhaps, Eddington mused, the missing mass was being transformed into energy in the form of light and heat.
Eddington quickly convinced himself and others that the amount of energy liberated by this transformation was enough to power the sun, at least in principle. The sticking point was how, exactly, four hydrogen atoms would come together to form a single helium atom. Atoms are electrically neutral, so they have no problem at all bumping into each other -- all the more so in the hot interior of the sun, which Eddington calculated to be 40 million degrees Kelvin. However, in order to fuse four hydrogen atoms into a helium atom, it is the atomic nuclei that have to come into contact, and those are positively charged and instantly repel each other. If the hydrogen atoms are moving fast enough, the repulsion can be overcome and the atoms will fuse, but there is a catch: in order to fuse more than a trivial amount of hydrogen, the temperature had to be much higher than Eddington calculated -- more like tens of billions of degrees Kelvin. Such a hot sun is incompatible with the sun's current size; at those temperatures, the sun should be much more bloated than it is. Eddington was convinced therefore that hydrogen was somehow fusing instead at the much colder temperature of 40 million degrees; he wrote in 1927,
We do not argue with the critic who urges that the stars are not hot enough for this process; we tell him to go and find a hotter place.
However, physicists could see no way around the electric repulsion problem.
As it turned out, however, there was a way. Just about the time that Eddington was fighting electric repulsion, the young Russian physicist George Gamov (1904-1968) showed that quantum mechanics explained how atomic nuclei could split apart in nuclear fission, even though the nuclear force holding the nucleus together was technically too strong to allow this to happen. The Welsh astronomer Robert Atkinson (1898-1982) and German physicist Fritz Houtermans (1903-1966) determined that the same mechanism could permit hydrogen nuclei to come together, even though the electromagnetic force holding them apart was technically too strong to allow this to happen, and they wrote this up in a paper in 1929. At last, the primary energy source of the stars was understood.
Let us recap, then. Stars derive their energy from fusion. In the process of fusing into a single atom of helium, four hydrogen atoms lose some mass; this mass is transformed into energy. The hydrogen atoms do not all come together at once -- even quantum mechanics does not allow this in the sun. Instead, the sun fuses hydrogen in two main ways:
1. In one process, a carbon atom "swallows" the four hydrogen atoms, one by one, emitting bits of energy along the way and becoming in turn different forms of nitrogen and oxygen, until it "burps out" the helium at the end and returns to its original carbon self. This is known as the carbon cycle.
2. In the second process, two hydrogens come together to form deuterium, a heavy form of hydrogen weighing about twice as much; this then collides with another hydrogen to form helium-3, a lighter form of helium which weighs -- you guessed it -- three times as much as ordinary hydrogen; finally, two helium-3 atoms collide to yield an ordinary helium atom and two hydrogen atoms. Those two hydrogen atoms are then available to be fused. This is known as the proton-proton chain.
Which process is more dominant in any given star depends on that star's temperature. For our own star and other cooler stars, the proton-proton chain is dominant. Stars much hotter than our own produce most of their energy via the carbon cycle.
The advent of computers made it possible to simulate stars of different mass and watch their progress in reasonable times (by human standards). Several billion years of stellar evolution could be compressed at first into weeks, then days, and then hours. Out of these simulations came the conclusion that more massive stars burn their hydrogen fuel much more quickly than less massive ones. Even though more massive stars have more hydrogen to begin with, they also burn hotter, and go through their hydrogen supply that much faster. The sun has enough hydrogen for about 11 billion years of uninterrupted fusion. Much larger stars -- say, those of about 50 times the sun's mass -- run out of hydrogen in only a few million years. On the other hand, much smaller stars, perhaps as small as a tenth of the sun's mass, run out of their hydrogen supply only after trillions of years. All of the stars this small are still burning steadily, because there hasn't been enough time since the universe was born for them to burn out.
But what happens after the hydrogen runs out? What then? Do they just fizzle out and gradually cool down, as many astronomers suspected? Or does something more fantastic happen? When the earliest stellar models were run through the computers, the simulations didn't yield anything like what we believed to be aging stars. Nor did they yield any kind of nonsensical result. They refused to yield anything at all; once the fusible hydrogen was depleted, the computers were unable to proceed further. There was hydrogen further out in the star, but only where it was so cool that fusion couldn't take place. It would take a completely new simulation technique to carry the life story of stars forward into death.
In my next essay, then, I'll talk about what happens when stars run out of hydrogen.
Adapted from Astronomical Games, September 2000.