Youth Who Left Ozark Mountains to Study Stars Causes Einstein to Change His Mind.
—Springfield Daily News, February 5, 1931
In 1925, with the Roaring Twenties half over, John T. Scopes was convicted of violating a Tennessee law against teaching evolution, or any theory denying that the universe had been created as described in the Book of Genesis. Biblical fundamentalists, offended by the suggestion that they were related to monkeys, might have been even more disturbed had they known about recent developments in astronomy. After more than 2,000 years of measuring, scientists were finding little reason to maintain the belief that there was anything special about the position of the sun within the Milky Way, or of the Milky Way itself within the endless sea of galaxies called the universe.
Had there been, say, a Hubble Star Trial resembling the Scopes “Monkey Trial,” one can imagine how the prosecution might have argued its case against the propounder of so great a heresy. It would have been difficult, and very unconvincing, to challenge the simple rules of geometry used to triangulate distances within the solar system or even to the nearest stars. Whether the baseline was the diameter of the Earth or the diameter of its orbit around the sun, the reasoning behind the measurements appealed to simple trigonometry and old-fashioned common sense.
More suspect, from the point of view of the astronomical fundamentalists, might have been some of the more indirect techniques—“mathematical mumbo jumbo,” a William Jennings Bryan might have objected. And he might have really gone to town with the Cepheids themselves. Perhaps the intrinsic brightness of the variable stars twinkling in the Magellanic Clouds is indeed signaled by how fast they pulse. But is it not a leap of faith to assume that the very same rule applies to Cepheids throughout the universe? Could not the good Lord have made his stars blink in any way he liked?
Here a Clarence Darrow, rising to the defense, might have prodded his client Hubble to remind the jurors how the distances derived from Cepheids harmonized so nicely with measurements made using other techniques, like the average luminosity of a galaxy’s brightest stars. Completely independent gauges all seemed to point toward the same conclusion. “But,” Bryan might have countered, “do not all these methods assume that stars near the Earth are fundamentally the same as stars in the far reaches of the heavens? Is that not a leap of faith?”
Darrow would have been smart to concede the point. What astronomers were taking on faith was the principle of uniformity—that the laws of physics apply equally in all parts of the cosmic realm. It would be an insult to the Creator to think he would design a universe in any other way.
Taking this idea to heart, one could now estimate the distance to any galaxy for which it was possible to pick out individual stars. Assuming that they were similar in kind to those nearby, you could guess their intrinsic brightness and use them as standard candles. Reaching into the astronomical grab bag of variables, novae, and so forth, you could piece together a map of the universe, or at least those regions closer in.
Beyond a certain distance, the method began to falter. Even with Mount Wilson’s 100-inch telescope, most nebulae by far were featureless blurs. There was little hope of finding a blinking Cepheid or even an exploding star. For very rough measurements, you could choose a closer galaxy whose distance you had already established and take the whole thing as a standard candle: estimate its intrinsic brightness and assume that the farther galaxies produce approximately the same amount of light. Then the inverse square law would kick in. The method was similar to that used in the eighteenth century when astronomers assumed, both out of ignorance and for convenience, that all stars were equally bright, making dimness a simple gauge of distance—and it was just as prone to error. Maybe the galaxy you were using as your standard was atypical, far brighter or dimmer than most. Still, if you averaged together the brightness of several known galaxies and used that as your yardstick, you could at least make a plausible argument. With this crude method, astronomers were reaching beyond Andromeda, finding galaxies whose light took millions of light-years to reach their telescopes.
The process was a bit like constructing a tall building with each level resting on the one below. With the diameter of the Earth as a baseline, you measure the distance to the sun. Assuming that figure is correct, you know the width of Earth’s orbit, a larger baseline from which to triangulate the very closest stars. Building on this information, you chart the sun’s own motion through the galaxy, providing an enormous baseline that, with some statistical sleight of hand, lets you measure the Cepheids and calibrate Henrietta Leavitt’s yardstick, and from there you make the next great leap.
The higher you climb, the more precarious the structure becomes. Perched with their telescopes at the loftiest level, astronomers knew it was foolish to be overly confident. At any moment a lower support might buckle. Everything they had built could come tumbling down.
2
Compared with the rickety nature of the distance scales, the celestial speed stick was fairly robust. Because of the Doppler effect, anything that emitted light (stars, galaxies, clouds of gas) could be clocked according to its color shift—a shrieking high-pitched blue if it was speeding toward you, a moaning low-pitched red if it was speeding away.
More specifically, the velocity was determined by the displacement of an object’s spectral lines. The method depended on a discovery by the German scientists Gustav Kirchhoff and Robert Bunsen, who found in the 1850s that they could identify chemical elements by burning them in a flame and refracting the glow through a prism. Certain colors would stand out in the spectrum—a combination of bright vertical lines as unique as a fingerprint. Sodium, for example, burns yellow. Seen through a spectroscope, it can be identified by a pair of bright “emission lines” at precise points in the yellow part of the spectrum.
When Kirchhoff and Bunsen made the discovery, the existence of atoms was still controversial. Once they were discovered, the effect could be simply understood: when an atom is energized, its electrons jump into higher orbits. When they fall back down they emit various frequencies of light. Every kind of atom is built a little differently, its electrons arrayed in a specific way, resulting in a characteristic pattern.
For similar reasons, if you shine a light through a gaseous substance, like hydrogen or helium, certain colors will be filtered out. The result in this case is a characteristic pattern of black “absorption” lines interrupting the spectrum—another unique chemical fingerprint. (The same colors marked by the absorption lines would appear as bright emission lines if the element was burned.) A scientist named Joseph von Fraunhofer had shown that lines like these appear in the spectrum of the sun. Using nothing more than a prism, one could stand on Earth and decipher the composition of the glowing orb 93 million miles away.
The natural next step was to add prisms to telescopes and analyze the chemical composition of stars and nebulae. They too exhibited the dark Fraunhofer lines, but not in the expected positions. They were displaced toward the red or blue end of the spectrum. Assuming this was caused by the Doppler effect, you could precisely gauge a galaxy’s velocity. A few, like Andromeda, were blue-shifted, approaching the Milky Way, but these were exceptions. Most were severely red-shifted, hurtling away at extremely high speeds.
As the 1920s drew toward a close, astronomers were finding hints of an even stranger phenomenon: the smaller, fainter, and thus presumably more distant a nebula was from Earth, the greater its redshift. Some astronomers dismissed this as an anomaly, a flaw in their techniques. Farther objects were obviously harder to analyze than closer ones. Some kind of systematic error might cause observers to overestimate the more distant redshifts. To correct for the mistake, they added a fudge factor to their calculations, a term they labeled K. It was considered, at this point, no more than a Band-Aid on the equations. When observations improved, it could be peeled off and thrown away.
There was also a more interesting possibility: that the K term was describing a real physical effect—that redshift actually did increase with distance. Grasping for an explanation, some theorists proposed that they were witnessing a heretofore unknown quality of light: the farther it traveled, the more its waves stretched and sagged toward the red end of the spectrum. This became known as the “tired light” theory. Perhaps, some proposed, the cause was some Einsteinian peculiarity of curved space-time.
Finally, it was possible that distant galaxies truly were flying away faster than closer ones. This seemed almost too good to be true. In a universe like this, redshift would provide the ultimate measuring stick. The distance to anything, no matter how far, could be measured as long as you could gather its light. Assuming that all stars are made from the same basic ingredients—hydrogen, helium, et cetera—they can be expected to exhibit the familiar patterns of spectral lines. The more the lines appear to be displaced, the faster the galaxy is moving, and—if the theory is correct—the farther away it is.
During a tour of Europe in 1928, Edwin Hubble, now the toast of his profession, heard reports of this curious redshift-distance connection. When he returned, he decided to investigate. He asked an assistant, Milton Humason, to train the 100-inch Mount Wilson telescope on some distant nebulae and see how their spectra behaved.
Even more than Henrietta Leavitt, Humason seemed an unlikely candidate for the astronomical hall of fame. He began his career at Mount Wilson as a mule driver, carrying material and supplies up the mountain. He married the daughter of the observatory engineer and talked his way into a job as janitor. Given the chance to learn how to make photographic plates, he quickly proved himself to be a very good photographer of starlight and was promoted to assistant astronomer. He had a grade school education.
Humason began his assignment by targeting a nebula so far away that no one had been able to measure its redshift: NGC 7619. The light, funneled through a prism, left its rainbow on the photographic plate. When it was developed it showed two familiar dark lines, indicating the presence of the element calcium. As expected, the lines were pushed toward the red. What was unexpected was how very big this displacement seemed to be. Humason took another picture to confirm the result. Working with a Mount Wilson computer (referred to in his paper simply as Miss MacCormack), he reported that he had clocked a galaxy speeding away at a rate “twice as large as any hitherto observed”: 3,779 kilometers per second, or more than 8 million miles per hour, a velocity that would get you from here to the moon in under two minutes.
In the following weeks, Humason measured more redshifts while Hubble scrutinized the results. By now he had compiled a list giving the recessional velocities of forty-six nebulae. He believed he had reliable distances to about half of them, derived from Cepheids, novae, and other yardsticks. For this sample, speed indeed seemed to increase with distance, and in a delightfully straightforward way. Some astronomers had suggested that the relationship might be “quadratic”: velocity would increase as the square of the distance. Others suggested more elaborate equations. What Hubble found could hardly have been simpler: A nebula twice as distant as another would be traveling at twice the speed. Triple the distance and the velocity would triple as well. The relationship was what a mathematician calls linear. Take any nebula in the universe and divide its speed by its distance. The result is always the same number—about 150 by Hubble’s reckoning.
This powerful number was none other than the mysterious K astronomers had been puzzling over. The “error” was apparently not an error at all but a factor describing how redshift increased with distance. A galaxy that was 1 million light-years from Earth was receding at about 150 kilometers per second. A galaxy 10 million light-years away traveled at 1,500 kilometers per second. Velocity equals distance times K. Or, more significantly, distance equals velocity divided by K. Applying his new formula to NGC 7619, the galaxy Humason had clocked at the breakneck speed of 3,779 kilometers per second, put it at more than 20 million light-years from earth.
It is impossible to sense from Hubble’s typically understated paper, or from the droning account he delivered a few years later in a lecture series called “The Realm of the Nebulae,” what he felt as the pieces of a new picture of the universe fell into place. Conservative as always, he cross-checked his measurements, testing whether they resulted in absurd conclusions. Using the galaxies’ apparent magnitudes and his new Doppler-derived distances, he computed how bright they would really be. The results were reassuring, comfortably within the range of galaxies closer by.
In the following months, Hubble and Humason continued to test the theory. Hubble would calculate a nebula’s distance using various measuring sticks and predict the redshift before Humason had even measured it. By now they were clocking galaxies with velocities as high as 20,000 kilometers per second (from Earth to moon in 20 seconds), putting them more than 100 million light-years away.
The numbers were so large that some astronomers were initially doubtful. On a visit to Pasadena, Harlow Shapley told a colleague, “I don’t believe these results.”
Not that Hubble or his assistant would have cared. Humason remembered one of his last encounters with Shapley, when he was still working on Mount Wilson. Scrutinizing plates of Andromeda with the blink comparator, Humason had spotted what he believed to be stars that varied periodically in brightness. This was more than two years before Hubble made his landmark discovery, establishing with Cepheids that Andromeda was a distant, neighboring galaxy. Humason marked off the places where these anomalies occurred and took the plates to Shapley.
Dismissively explaining why they couldn’t be Cepheids, Shapley took out his handkerchief and wiped the plates clean, erasing the data. A few months later he departed for Harvard.
3
At heart, Hubble, like Edward Pickering, was an observer not a theorist, leery of speculating beyond what his eyes could see. It took an Einstein to explain the theory and the mechanism of what astronomers were soon calling the Hubble shift (the K in the equations ceremoniously replaced by an H). Why were the galaxies moving—and why were they, with so few exceptions, all hurtling outward from the Milky Way?
Think of the galaxies as runners in a race. After a certain amount of time has passed, they will be distributed according to their swiftness, the fastest ones farthest from the starting line and the slowest ones closest in. But didn’t that imply that there was something special and outrageously non-Copernican about our position in the universe, as the point from which everything else was retreating?
A motionless universe was far easier to fathom, even at first for Einstein. When his own general theory of relativity implied that the cosmos may be expanding, he had rejiggered the equations, committing what he would come to regard as an embarrassing mistake. Now he knew that the adjustment had been superfluous. The universe really moved. Visiting Pasadena in 1931 he told Hubble’s wife that her husband’s work was “beautiful” and publicly conceded that his earlier conviction of a static universe was wrong. Hubble’s hometown newspaper in Missouri picked up the story: “Youth Who Left Ozark Mountains to Study Stars Causes Einstein to Change His Mind.”
Einstein was glad he could restore his theory to its pristine form. What his equations now described was a universe in which space itself is expanding. Second by second, the galaxies grow wider apart like dots on an inflating balloon. Viewed this way, Hubble’s discovery did not imply that the Earth was in a special position, at the center of the outward rush of everything. From any point in the cosmos the effect would be the same, with galaxies appearing to fly off in every direction. And if you could reverse the clock, everything would become closer and more compact, converging on a single point. The big bang. The universe had a beginning. And maybe it will have an end.
So rarefied a theory, now taken as gospel, was still of secondary interest to most astronomers. Hubble himself was noncommittal about the meaning of the Hubble constant and the Hubble shift. Expanding universe, tired light—it didn’t matter. What he knew for certain was that redshift, for whatever reason, increased with distance, and that gave him a way to measure as far as a telescope could see.