THE CEPHEID SCALE

By the end of the nineteenth century, astronomers had begun to recognize that the universe extended well beyond the solar system. The planets were far from Earth, but the stars were much farther. Still, they had no idea how far they really were. With the instruments of the day, parallax could be used to measure distances out to no more than a few tens of light-years.

Despite the Copernican revolution, astronomers still imagined Earth to be near the center of the universe. This was not just typical human self-centeredness. When astronomers counted stars, they found the number to fall off fairly uniformly in all directions from Earth; so it seemed we were near the center. They were unaware of the existence of interstellar gas that dims the light from farther objects equally in all directions and leaves the appearance of isotropy.

In 1908, Henrietta Leavitt was one of a small group of young, female “computers” working for Charles Pickering at Harvard College Observatory in Cambridge, Massachusetts. Pickering realized that professional astronomers were not needed to perform all the detailed work that was involved in poring over the many photographic plates exposed at the observatories of Harvard's several telescopes. In fact, young women who could be hired for a fraction of the wage did a very good job of analyzing what was on the plates and recording the exact positions, spectral classes, and brightness magnitudes of stars and other astronomical objects. Naturally, the women did no actual observing, given their “delicate natures.”

Pickering assigned Leavitt to look at the photos of variable stars—stars whose brightness magnitudes change over time. Harvard had an observatory in Peru, and Leavitt looked at photos taken there of a sample of stars within the Small Magellanic Cloud (SMC), which along with those in the Large Magellanic Cloud (LMC) are only visible in the Southern Hemisphere.

Comparing plates of sixteen giant stars called Cepheid variables, Leavitt made a discovery that would prove to be of huge significance. She noticed that the period of variation, that is, the time between peak luminosities, was longer when the star was brighter. Being at roughly the same distances from Earth, she reasoned that the observed luminosities of the Cepheids in the SMC should be directly related to their intrinsic luminosities. Leavitt had uncovered a connection between peak luminosity and period that would ultimately provide a means to measure far greater distances than is possible with parallax.

Leavitt's efforts were delayed by chronic illness. However, by 1912 she had gathered nine more Cepheids in the SMC and published a three-page article, No. 173, in the Harvard College Observatory Circular. The article contained a graph, on a logarithmic scale, of Cepheid brightness versus period and showed a clear relationship between the two. The graph was termed a period-luminosity relation.1

OFF CENTER

In 1908, the most powerful telescope of the time, a 60-inch reflector on Mount Wilson in California, built by George Ellery Hale, began looking up into the crystal-clear night skies above Los Angeles. In 1912, Harlow Shapley, a former crime reporter from rural Missouri who had gone on to earn a doctorate in astronomy from Princeton, joined the observatory staff. Shapley took an interest in “globular clusters,” spherical systems containing hundreds or even thousands of stars.

He (or his wife, Martha, who dabbled in astronomy) discovered that the clusters included many faint giant blue stars. Comparing their observed luminosities with the stars of that type near the sun, he estimated that they were at least fifty thousand light-years away—two orders of magnitude farther than the greatest distances of stars then known from parallax.

Unfortunately, globular clusters contain few or no Cepheids. Furthermore, Shapley found that some Cepheids had much shorter periods than the ones in the SMC used in Leavitt's sample. For this reason, at first Shapley hesitated to use Leavitt's period-luminosity relation on the Cepheid variables with shorter periods. However, he was not your typical cautious scientist and plowed ahead anyway. He used Leavitt's period-luminosity relation to measure the distances to those Cepheids he could see in his globular-cluster photos.

Here's how it works: You measure the observed period of a Cepheid variable. From the period-luminosity relation you obtain the intrinsic luminosity of the star. As the star radiates outward on the surface of an imaginary sphere of ever-increasing radius r, the energy per unit area will fall off with the sphere's area as 1/r² because of energy conservation. Thus, by comparing the measured luminosity with the apparent luminosity, you can determine the distance to the star.

For the clusters with no Cepheids, or those too faint to measure, Shapley estimated their distances using the brightest stars as distance markers. When the stars were not resolvable, he used the size of the cluster to judge distance.

By these methods, Shapley determined that the Milky Way is shaped like an ellipsoid and some three hundred thousand light-years across. He concluded that it must constitute the entire universe. Imaginative as he was, he could not imagine a larger universe.

Shapley noticed that his distribution of globular clusters was not centered on Earth. Rather the distribution appeared to be centered on the region of the constellation Sagittarius. (He was not the first to suggest this.) He estimated that the sun is sixty-five thousand light-years from the center.

Shapley's distances were in fact overestimates. Today we know the Milky Way is 100,000 to 120,000 light-years in diameter, with the sun 27,000 light-years from the center.

HIGH-SPEED ASTRONOMY

Percival Lowell was a descendent of the patrician Massachusetts family that first landed in Boston in 1639. He built an observatory in Arizona to pursue his obsession with Mars and the possibility that it contained artificial waterways or canals built by a past civilization on the red planet. In 1877, Italian astronomer Giovanni Schiaparelli (1835–1910) had reported seeing dark streaks across the surface of Mars, which he called canali, meaning “channels” (not “canals”). Unlike most astronomers, Lowell took them seriously and wrote three books on the subject that popularized the notion of life on Mars. The observatory was built on “Mars Hill,” which, at 7,250 feet, is three thousand feet higher than Mount Hamilton and fifteen hundred feet higher than Mount Wilson.

Lowell purchased a spectroscope that was an improvement over the one at Lick. However, Lowell had only a 24-inch refractor, which made it less suitable for spectral work.

Lowell hired the young Vesto Slipher right out of Indiana University in 1909. Slipher remained at the observatory until retiring in 1954, serving as director for thirty-eight years.

With great patience and skill, Slipher significantly improved the tricky instrument and pretty much did whatever Lowell asked, which was mainly planetary astronomy. However, in 1909, Lowell directed Slipher to take a spectrogram of what he called a “white nebula,” by which he meant a spiral nebula. Slipher did not get around to it until 1912, when he began a series of exposures of Andromeda, the largest spiral nebula in the sky, on four different plates. Then, in January 1913 he obtained his result: The spectrum of Andromeda was blueshifted, that is, shifted to shorter wavelengths. Assuming the mechanism was a Doppler shift, Slipher calculated that Andromeda is moving toward us with a speed or radial velocity of 300 kilometers per second.2 He wasn't far off. Currently astronomers measure 301.

At the time, this was the highest velocity ever measured in nature. The radial velocity of Andromeda is ten times the speed of Earth around the sun, thirty kilometers per second, which is about the same as the typical speed of a star in the Milky Way.

Slipher's result was extraordinary, and other astronomers, especially those at the nearby competitor, Lick, were properly skeptical. Undaunted, Slipher continued his measurements, which were that much tougher for smaller nebulae. He found the spectrum of M87 (Messier catalog) was redshifted, indicating it is moving away from Earth with a radial velocity of one thousand kilometers per second—three times faster than M31, Andromeda, which is moving toward us and will someday merge with the Milky Way. By the summer of 1914, Slipher had measured the velocities of fourteen spiral nebulae and found most to be receding. In August, he gave a report to the meeting of the American Astronomical Society at Northwestern University. A tall, handsome young man just elected to the society named Edwin Hubble (1889–1953) was in the audience.

Slipher reported the spiral nebula move at an average speed of twenty-five times the average speed of stars in our galaxy. He received a standing ovation and congratulations from his colleagues at Lick. However, the evidence that the spirals were separate island universes was still not conclusive. There had to be a way to measure their distances.³

As mentioned in chapter 4, today astronomers define a quantity z called the redshift that equals the fractional amount by which an observed wavelength of spectral lines exceeds the wavelengths as measured in the laboratory. If it is negative, we have a blueshift. From the Doppler effect, z = v/c where v is the recessional velocity. This only applies for v << c; a more-complicated formula must be used at relativistic speeds.

A SPIRALING DEBATE

As we saw in chapter 3, near the end of the nineteenth century, James Keeler had taken beautiful photographs of hundreds of spiral nebulae with the Crossley telescope on Mount Hamilton. Nine years after Keeler's death in 1900, Heber Curtis (1872–1942) resumed Keeler's pursuit with the Crossley. By 1913, he had assembled two hundred photos of spirals.

On July 19, 1917, George Richey (1864–1945), observing with the 60-inch reflector on Mount Wilson, three hundred miles from Mount Hamilton, took a long-exposure photo of the spiral nebula NGC 6946 (NGC designates The New General Catalogue, compiled in 1888). Comparing with three earlier pictures, Richey saw a pinpoint of light hear the edge. He concluded it was a nova, one of the points of light that occasionally appear in the sky and quickly fade away.

Curtis noticed the phenomenon in three different nebulae. Searching through plates, he found separate novae had appeared in the spiral NGC 4321 in 1901 and 1914. It seemed very strange to him to have so many novae appear in spiral nebulae.

Furthermore, Curtis noticed that some nebulae occasionally, though rarely, exhibited relatively bright bursts. For example, such bursts were seen in Andromeda in 1885 and in the constellation Centaurus in 1895. Curtis placed these bursts in another class, which are now called supernovae.

Intrigued, other astronomers began searching their plates. Continuing his own search, Curtis observed that most of the novae in spirals resembled novae elsewhere but were much dimmer. To account for their faintness, he concluded that they had to be millions of light-years away.

Curtis became the champion of the notion that spiral nebulae were “island universes,” galaxies of stars located far beyond our own Milky Way. But most astronomers remained skeptical. In the meantime, World War I intervened and most astronomers, including Curtis (but not Shapley), joined the war effort.⁴

After the war, the spiral controversy continued, with Harlow Shapley holding strong to his view that they were within our galaxy. There he appeared to have strong support from the Dutch-born astronomer Adriaan van Maanen, who worked on Mount Wilson. Van Maanen, who had a good reputation as a meticulous researcher, claimed to have measured rotational periods for spiral nebula that implied that if they were separate galaxies as large as the Milky Way, then their spiral arms moved faster than the speed of light. His results were purportedly corroborated by other observers on Mount Wilson, Lowell Observatory, and in Russia and the Netherlands.5

On April 26, 1920, Shapley and Curtis engaged in what is called the “Great Debate” at an evening public meeting during a three-day National Academy of Sciences convention in Washington, DC. The day before, the Washington Post announced that “Dr. Harlow Shapley, of the Mount Wilson solar observatory, will discuss evidence which seems to indicate the scale of the [Milky Way] to be many times greater than is held…. Dr. Heber D. Curtis, of the Lick Observatory, will defend the old [my emphasis] theory that there are possibly numerous universes similar to our own, each of which may contain three billion stars.”⁶

If we are to judge from a newspaper article, which is always a risky business, the general tenor at the time may have been that the “old” idea that there were other galaxies besides our own was being superseded by Shapley's more recent discoveries.

Actually, it wasn't much of a debate. Rather there were two back-to-back lectures that really did not address each other, with any rebuttals left to the discussion afterward.

Shapley spent most of the debate describing his estimate of the size of our galaxy, which was already ten times what had been earlier thought. He aimed at a public audience and gave few technical details. Curtis gave a more technical talk, focusing on spiral nebulae, emphasizing the Andromeda nova, which was too bright to be within our galaxy, and the high speeds of the spirals.

Curtis also questioned the size estimate for the Milky Way, saying it was ten times smaller than Shapley estimated. Here he was three times too small while Shapley was three times too big, so we can split the difference.

Curtis had to admit that if the spiral nebulae were separate galaxies and one reasonably assumed they were on the same order of magnitude of size as the Milky Way, they would have to be at least three hundred thousand light-years away.

No good record of the actual debate survives. But the two astronomers, competent scientists that they each were, agreed to write articles on their positions for the Bulletin of the National Research Council, which appeared a year later apparently substantially modified from what they presented orally after they exchanged a series of drafts.7

The public began to take notice. On May 31, 1921, the New York Times reported on its front page that Shapley had proved that “the little speck of light around which a tiny shadow called the earth revolves, is 60,000 light-years from the centre of the universe.” Shapley had replaced Pickering as director of the Harvard Observatory that year. He is quoted in the article as saying, “Personally, I am glad to see man sink into such physical nothingness, and it is wholesome for human beings to realize of what small importance they are in comparison with the universe.”⁸

In 1925, the Royal Swedish Academy contacted Harvard Observatory about the possibility of awarding the Nobel Prize for Physics to Henrietta Leavitt. They were unaware that she had died four years earlier, on December 12, 1921, of stomach cancer at the age of fifty-three. The prize is not awarded posthumously.

THE REALM OF HUBBLE

On September 11, 1919, thanks to the tireless efforts of George Ellery Hale, a 100-inch reflecting telescope came into operation on Mount Wilson. A week later, Edwin Hubble joined the staff after serving as a captain in the army during the Great War, where he saw no combat. In 1923, Hubble was using both the 60-inch and 100-inch to survey nebulae, and he turned his attention to NGC 6822, a nebula in Sagittarius that resembles the LMC.

Hubble found five variable stars in NGC 6822 and asked Shapley in Harvard if he could look at the object in his store of plates. Shapley had used the observed magnitudes of the brightest stars to estimate that NGC 6822 is about a million light-years away. He admitted that it was “probably quite beyond the limits of the galactic system,” but it wasn't a spiral so Shapley continued to insist that spirals were “neither galactic in size nor stellar in composition.”9

Hubble found eleven Cepheids in NGC 6822 and used them to obtain a distance of seven hundred thousand light-years. However, the great discovery that would bring him world fame occurred in early 1924 when he found a Cepheid variable in Andromeda and determined it to be nine hundred thousand light-years away. Hubble wrote Shapley, who told a colleague when he received the letter: “Here is the letter that destroyed my universe.”¹⁰

Actually, in 1922 the Estonian astronomer Ernst Öpik had published a more-accurate measurement of the distance to Andromeda by means of a novel technique in which he used the rotational velocity of a galaxy, which depends on the mass of the galaxy, and assumed its luminosity is proportional to its mass. His estimate was 1.5 million light-years, compared to Hubble's estimate of nine hundred thousand light-years. This was closer to the modern value of 2.5 million light-years.

Nevertheless, Shapley played devil's advocate for a while, which was a perfectly proper position to take as the prime expert supporting the opposing theory. When new discoveries are made in science, they are not so obvious at the time to those in the trenches, and Hubble was also properly being very cautious and conservative.

In the midst of it all, Hubble married the daughter of a wealthy Los Angeles banker, and the two embarked on a three-month honeymoon that included a tour of Europe.

After returning, Hubble turned to other nebula, in particular the beautiful face-on spiral M33 in the Triangulum constellation. There he found twenty-two Cepheids implying again a distance of at least a million light-years. Measuring the periods of all these variable stars in those days before computers or even hand calculators was tedious and laborious. And Hubble was bothered by the measurements on spirals made by his senior colleague at Mount Wilson, van Maanen. If they were correct, the spirals could not be extragalactic, so he was reluctant to announce his results publicly or express his own doubts about van Maanen's conclusions.

However, Hubble's results quickly made it around the world by what was already a remarkably efficient astronomical grapevine (it's instantaneous today). Even the New York Times got wind of it, and on November 23, 1924, it headlined, “Dr. Hubbell [sic] Confirms View That There Are ‘Island Universes’ Similar To Our Own.”11

Despite Hubble's hesitation, the astronomical community took his results seriously because they were based on what was by then well-established, the Cepheid-distance scale. Van Maanen was assumed to be wrong, and eventually flaws in his methods were uncovered. Hubble shared a $1,000 prize from the American Association for the Advancement of Science with parasitologist Lemuel Cleveland for his discovery of protozoa in the digestive tracks of termites.¹² Hubble published his results in 1925 in the Publications of the American Astronomical Society.¹³

Shapley regretted his move to Harvard. He believed he would have made the same discoveries as Hubble had he remained at Mount Wilson. But, in the end he was gracious, saying Hubble earned his fame and was “an excellent observer, better than I.”¹⁴

Hubble, though, regarded van Maanen's findings to be a stain on his great accomplishment, nursing a personal grudge as they continued to work on the same mountain. Van Maanen only reluctantly admitted he may have made some errors and promised to follow up. He never did.¹⁵