The Day We Found the Universe (21 page)

A globular cluster appears through a telescope as an assembly of brilliant specks of light hovering around a dense and blazing core. With stars packed in like subway commuters at rush hour, the cluster offers a far more exotic celestial environment than our local stellar neighborhood. Alpha Centauri, the star closest to the Sun, is some 4 lightyears away. But if the Sun were in the center of a packed globular cluster, it would have thousands of stars closer than that, covering Earth's sky like a sequined blanket visible both day and night. Near misses between stars would be commonplace.

That a globular cluster is a highly spherical collection of stars was not known until the 1600s, with the advent of the telescope. Before that, ancient astronomers simply noted the objects on their sky charts as a “lucid spot” or a lone “hazy star.” Today, these clusters are known to be arranged as a globelike halo, surrounding the disk of the Milky Way somewhat like bees buzzing around a hive. But as late as the 1910s, when Shapley began his observations, astronomers didn't know that, nor exactly how big an individual globular cluster was. Some even pondered if they were island universes in their own right. Shapley himself believed that was true when he was starting out: “It is quite obvious that a globular cluster … is in itself a stellar system on a great scale—a stellar unit which without doubt must be comparable to our own galactic system in many ways,” he wrote in the first paper of his study. Some dabbled with the idea that a spiral nebula was an early stage of a globular cluster about to form: Like an open flower closing at twilight, the spiral over time would fold up into a ball. Shapley's goal was to learn the globulars' true sizes, distances, and compositions and see if such ideas were valid.

Shapley's initial observations were fairly basic. Using the 60-inch telescope, he simply surveyed the colors and magnitudes of the stars in the most prominent clusters. These included Omega Centauri (the biggest of them all), the Hercules cluster, and M3, a globular noted by Charles Messier in 1764. Shapley had no idea where this would lead, but that was standard practice in astronomy: Gather as much data as you can when faced with the unknown and keep your eye out for unusual trends. If anything, Shapley hoped his observations might help Hale in his quest to understand how stars aged and evolved, still quite a mystery to astronomers in the early twentieth century.

Globular Cluster M80
(The Hubble Heritage Team
[
AURA/STScI/NASA
])

As his collection of photographs mushroomed, though, Shapley began to identify Cepheids which he knew would serve as his measuring tape out to the globular clusters. He was quite aware of the paper that Henrietta Leavitt had published just a couple of years earlier and intended to apply it. “Her discovery … is destined to be one of the most significant results of stellar astronomy,” Shapley later wrote to her boss, Pickering.

What was needed was a reliable distance to a Cepheid—any Cepheid, anywhere in the sky—that could serve as the calibration for determining the distances to all other Cepheids using Leavitt's period-luminosity law. That was the beauty of her discovery: Know the distance to just one Cepheid and you know the rest.

Distance measurements have long been a problem for astronomers. To our eye, the celestial sky resembles a dark bowl with pinpoints of light affixed to it—everything appears to be the same distance away. But in reality the stars we see reside at vastly different ranges. Bluish-white Sirius, the brightest star in the heavens, is located 8.6 lightyears from Earth; Vega, the prominent summertime star in the constellation Lyra, lies 25 lightyears away. How do astronomers arrive at these numbers? “Parallax” is one surveying technique. Parallax is the apparent change in a star's position on the sky when observed first at one end of Earth's orbit and then six months later at the other end (similar to the way an object close by will appear to shift when you view it first with one eye, then the other). By setting the radius of Earth's orbit as a baseline and knowing the angle of shift in the star's parallax, a bit of geometric triangulation determines the star's distance from the Sun directly. Astronomers devised the term
parsec
to describe the distance between Earth and a celestial object that displays a
parallax
of one arc
sec
ond of angular measurement on the sky. (One parsec equals 3.26 lightyears.) The parallax method is useful out to several hundred lightyears. After that, the change in a star's position is too small to be discernible by ground-based telescopes, which is why Leavitt's law was so treasured. It would enable astronomers to extend their distance surveys much farther outward. It would have been nice if a Cepheid resided fairly close to our Sun; then astronomers could have measured the star's parallax and gotten their calibration fairly easily. Unfortunately, there was no Cepheid within reach of a direct parallax measurement from Earth in Shapley's day. Nature was not so accommodating to astronomers. (The closest Cepheid to us is Polaris, the North Star, located about 430 lightyears away. Polaris is actually a three-star system, one of which is a large yellow Cepheid that completes its dim/bright cycle every four days.)

The first person to try to confront the Cepheid distance problem was Ejnar Hertzsprung, who had initially recognized that Leavitt's twenty-five variables in the Small Magellanic Cloud were specifically Cepheid stars. He began to look at the Cepheids best studied within the Milky Way, thirteen in all. He couldn't measure their parallax (they were too far away), but he could consult a chronological sequence of astronomical atlases to see how far the stars had moved across the sky, at right angles to our line of sight, in their travels through the Milky Way. It was a matter of determining how their celestial coordinates had changed over the years. Astronomers refer to this advance as a star's “proper motion.” From another type of catalog he looked up how fast they were moving either toward or away from Earth based on the stars' blueshifts or redshifts (a rough gauge of their overall velocity). In an imaginative leap, he then estimated the Cepheid's distance by comparing the star's measured velocity with how fast it
appears
to be moving across the sky from our far away vantage point. The more distant the star, the slower it seems to journey across the sky. (His actual mathematical procedure, which also involved the Sun's motion through the galaxy, was more complex, but this provides the basic idea.) Hertzsprung's approach in the end provided a crude statistical calibration, one that he then applied to Leavitt's Cepheids in the Small Magellanic Cloud. He concluded that the cloud was 30,000 lightyears distant, one of the greatest distances then measured for a celestial object. This demonstrated for the first time the potential power of Leavitt's discovery.

“I had not thought of making the very pretty use you make of Miss Leavitt's discovery,” Henry Norris Russell wrote Hertzsprung when this result came out. Russell had employed a similar technique around the same time, but his aim was to determine the average magnitude of a Cepheid. In the process, he concluded that they were giant stars, far bigger than our Sun. Inspired by Hertzsprung, Russell proceeded to make his own distance calculation to the Small Magellanic Cloud, arriving at 80,000 lightyears. Both estimates were highly uncertain and turned out to be far less than current distance measurements (210,000 lightyears), but each figure was still astoundingly huge for its day.

Shapley soon adopted Hertzsprung's approach, although he used only eleven of Hertzsprung's thirteen Cepheids in his calibration, suspecting that two of them were peculiar. Just like Hertzsprung, he counted on a simple rule of perspective: The farther away a moving object is located from you, the slower it will appear to travel. A far-off plane seems to crawl along the sky, while a plane closer in going at the same speed would zoom right past you. After estimating an average velocity for a star, Shapley checked how his eleven Cepheids were journeying across the sky. The slower the apparent velocity, the more distant the Cepheid.

It was at this point, though, that Shapley parted company with Hertzsprung. He didn't use Leavitt's period-luminosity relationship, which was based solely on stars in the Small Magellanic Cloud, but instead constructed his own relationship based as well on the Cepheids in the Milky Way, in order to obtain an “improved and extended” period-luminosity law combining both sets of variable stars. He then applied his new rule to the Cepheids found in the globular clusters. He would monitor a Cepheid to peg its period and then calculate the star's distance from his graph.

This worked as long as Shapley could find Cepheids in his globular clusters. Some of the clusters had none at all, as far as he could observe. What they did harbor were variables that were not quite the same. These variables changed quite rapidly, in a matter of hours rather than days or months. There was no guarantee that they behaved in the same way as Leavitt's Cepheids.

Shapley tried mightily to check with Leavitt on this question, writing several times to her boss, Edward Pickering, on whether she had detected fast variables in the Magellanic Clouds and found them to obey her rule. Pickering assured him that photographs were being taken. But progress on the question was occurring at a glacial pace. Pickering was keeping Leavitt busy with work he considered more important. “Routine stuff,” decried Russell to Shapley at one point. “I fear, however, that I am not the man who may justly raise my voice in criticism.”

Eager to move forward, Shapley simply decided to treat his fast variables as if they did follow Leavitt's rule. He extended the Cepheids' period-luminosity relationship to include all these variables, both slow and fast. “This proposition scarcely needs proof,” he had boldly asserted in one early paper, though it was a very controversial decision. But by doing this, Shapley was able to determine the distances to the nearest globular clusters—a formidable task, as the stars were very faint. For clusters farther out, too remote to spot any variables, he resorted to using the brightest stars as distance markers. He just assumed that the brightest stars in a distant globular cluster had a similar magnitude on average to the brightest stars in a nearby cluster. And when the stars themselves could no longer be adequately resolved, he judged distance by the apparent size of the globular cluster in the sky. “The whole line of reasoning… was brilliant,” concluded astronomer Allan Sandage in a review of this technique decades later. Bailey, at Harvard, could have carried out this effort before Shapley, but he was overly cautious about the variables. To him there were too many uncertainties about their nature, so “definite conclusions from these data cannot be safely made,” he reported. Shapley had no such qualms.

But it was yeoman's work, painstaking routines that took four years to complete. Shapley was securing the distance to every Milky Way globular cluster known at the time, sixty-nine in all. With the assistance of Edison Hoge, he took some three hundred photographs. Some exposures were only ten seconds in length, but others lasted up to two hours. Most took minutes. Afterward there was the brutal labor at the work-table analyzing what the images revealed. By 1917 he was writing a colleague that “the work on clusters goes on monotonously—monotonous as far as labor is concerned, but the results are continual pleasure. Give me time enough and I shall get something out of the problem yet.” By then the war was on, but Shapley didn't sign up. He claimed that Hale had convinced him to stay at his job.

Some of the globular clusters (circled) surrounding the Milky Way
(Harvard College Observatory, courtesy of AIP Emilio Segrè Visual Archives)

Shapley didn't particularly enjoy his nights alone with the stars. What drove him back to the telescope month after month were his findings. With the first hint of dawn in the east, as the dome slit slowly closed with a noisome squeal, nearby coyotes would answer in kind with a serenade of high-pitched howls. At night's end, he and the other astronomers would walk back to the Monastery, sometimes whistling a merry tune if the viewing went well and forgetting that they might be disturbing the daytime observers—the solar astronomers—who were still fast asleep. Once in bed themselves, though, the nighttime observers could easily be wakened by the stirrings of the daytime crew. Both sides were together at noontime lunch, which offered the opportunity to settle any squabbles.