—Harlow Shapley, writing about Henrietta Swan Leavitt
For two boys hailing from rural Missouri, Harlow Shapley and Edwin Hubble didn’t have much in common, except perhaps the size of their egos. Shapley had been born in 1885 on a hay farm near the Ozarks and dropped out of school to become a police beat reporter for a small-town newspaper. He had completed the equivalent of the fifth grade. Only a few years later did he earn a high school diploma and enroll at the University of Missouri. There he was diverted from journalism into astronomy, finally moving on to Princeton to study under Russell, who once told Edward Pickering, “He is the best student I ever had.” For all Shapley’s self-assured swagger—he was certain that he had mapped the length and breadth of the universe—he never quite lost his rough country edge.
Four years later, and about seventy miles east of the Shapley farm, Hubble was born. The family was more prosperous than the Shapleys (Mr. Hubble was an attorney turned insurance executive). A straight-A student and an athlete, Edwin won a scholarship to the University of Chicago and became a Rhodes Scholar at Oxford, where he studied law and picked up the fake British accent that would grate on some of his colleagues like a kid squeaking a balloon.
Edwin Hubble (Courtesy of the Archives,
California Institute of Technology)
Hubble, like his father, didn’t take to practicing law. After briefly working as a high school teacher in Indiana (appearing before his classes in knickers and a cape), he returned to Chicago to pursue a doctorate in astronomy. Then, after serving as an officer in World War I, he came to Mount Wilson, where he and Shapley found themselves working uncomfortably under the same dome-shaped roof. What, Shapley wondered, was a Missouri boy doing exclaiming, “Bah Jove,” or remarking that a plan had “come a cropper”? Taking note of Hubble’s aristocratic and somewhat military bearing, some astronomers began referring to him as the Major.
Hubble, a rather reserved sort, found Shapley overbearing and erratic—shooting off one wild idea after the other—and he was particularly put off by Shapley’s friend, the garrulous Dutch astronomer Adriaan van Maanen, whose love for dinner parties and socializing made him something of a standout in stodgy Pasadena. Van Maanen was also known as a meticulous astronomer—his measurements of the rotational velocity of the spiral nebulae provided the strongest argument against island universes. Having rechecked and refined his data, he remained adamant: either the spirals were small and nearby or they were spinning at insanely high speeds. He believed, as his teacher Kapteyn had taught him, that there can be only one galaxy, the Milky Way.
Hubble himself leaned toward the island universe theory, but for now he wasn’t involving himself in the controversy. He hadn’t come to Mount Wilson to confer with mortals about their astronomical opinions. The answers would be found in the stars. Within weeks he was sitting through long nights beneath the observatory dome, consorting with the sky. He spent Christmas Eve 1919 with his eye to Mount Wilson’s new 100-inch telescope, 40 inches greater in diameter than the one Shapley had used to map out his Big Galaxy. For the next three decades it would be the largest in the world. Hubble was looking particularly hard at nebulae and wondering, perhaps, when Harlow Shapley was going to get off his mountain.
2
By now, Henrietta Leavitt and her widowed mother had taken up housekeeping in a recently built brick apartment building on Linnean Street and Massachusetts Avenue, several blocks from Harvard Observatory. Although she was largely occupied with more routine tasks, variable stars were still very much on her mind. In 1920 she wrote to Shapley seeking his advice. Where would he suggest she focus her research next? He replied, still hammering on an old theme, that it would be of “enormous importance in the present discussion of the distances of globular clusters and the size of the galactic system” if she would plot the periods of some of the dimmer variables in the Small Magellanic Cloud, those “just fainter than the faintest already studied.” This is what he had been pestering Pickering about until several months before his death.
And perhaps she would see if her discovery about the Cepheids also held for those in the Large Magellanic Cloud. “Does the period-luminosity law apply there?” He was treating her almost like a colleague. Soon he would be her boss.
Shapley had been overestimating how badly Harvard wanted him. At first the university’s president, Abbott Lawrence Lowell, had considered him the obvious choice to succeed Edward Pickering as director. But after consulting with several astronomers, Lowell found himself leaning toward offering Shapley the number-two spot, with an older, more experienced astronomer like Henry Norris Russell running the show. Shapley struck some of his colleagues as young and immature, and perhaps too brash for Harvard. “He is much more venturesome than other members of our staff,” Shapley’s boss, George Ellery Hale, confided to Lowell, “and more willing to base far-reaching conclusions on rather slender data.” And, as Shapley had feared, his lackluster performance at the Great Debate hadn’t helped matters. Even Russell came away persuaded that his protégé was not ready to run Harvard Observatory. “Shapley couldn’t swing the thing alone,” Russell told Hale. “I am convinced of that after trying to measure myself with the job, and observing Shapley in Washington. But he would make a bully second.”
In the end Shapley, willful as ever, got what he wanted. Ultimately Russell turned down Harvard’s offer and Shapley made it clear that he wouldn’t settle for anything less than the directorship. Hale interceded on his behalf and Harvard agreed to try out the young astronomer on a one-year probation. In the spring of 1921 he moved to Cambridge to take over where Pickering had left off.
“MARCH 28, 1921:Dr.Shapley arrived!”wrote Annie Cannon, who had become one of the most accomplished of the observatory’s women assistants, in her diary. “I like him. So young, so clean, so brilliant.” Like Henrietta Leavitt, Cannon was deaf, though only partially so. The following week she and a friend invited Shapley over for dinner and they all went to the symphony.
By now Leavitt was the head of stellar photometry, and the ebullient Cannon was curator of the photographic plate collection and chief compiler and overseer of the Henry Draper Catalogue of stellar spectra. Ultimately it filled nine volumes with more than 225,000 stars classified according to their spectral type, from the hottest white-blue stars to the cooler yellow, orange, and red. (Cannon’s categories were called, cryptically, O, B, A, F, G, K, and M, which astronomers, some of the men anyway, remember with this mnemonic: “O Be A Fine Girl, Kiss Me”).
Under Pickering, the status of the women computers had continued its slow climb. He even tried, with no success whatsoever, to persuade the president of Harvard to grant Cannon the prestige of an academic appointment, or at least to include her name in the school’s catalog. Women were praised, a little condescendingly, for being good at detail work, the numerical needlepoint of analyzing astronomical imagery, but deeper matters were still reserved for the men. One of the more overqualified assistants, Antonia Maury, a Vassar graduate, chafed under the tedium. “I always wanted to learn the calculus,” she later said, “but Professor Pickering did not wish it.”
Maury, truth be told, could be a pain to work with. She had been hired because she was Henry Draper’s niece. Her work was slow and erratic, prompting her aunt to apologize for her behavior. “I shall be happy,” she had written to Pickering, “when you are rid of the annoyance.” But Maury’s bad attitude was inflamed by a feeling that she was being discouraged from making original contributions.
In her own diary, Williamina Fleming, the housekeeper turned astronomical assistant, expressed the sense of frustration and stoicism some of the computers felt: “If one could only go on and on with original work, looking to new stars, variables, classifying spectra and studying their peculiarities and changes, life would be a most beautiful dream; but you come down to its realities when you have to put all that is most interesting to you aside, in order to use most of your available time preparing the work of others for publication. However, whatsoever thou puttest thy hand to, do it well.”
Cannon was happier with her lot. When a young British astronomer named Cecilia Payne arrived to study at the observatory in 1923, she wondered how Cannon could have spent all those years under Pickering meticulously classifying stars without speculating on what the new taxonomy might mean. In Cannon’s case, Payne came to conclude, theorizing was against her nature: “She was a pure observer, she did not interpret.” And she seemed to rely less on reason than on instinct. “She was like a person with a phenomenal memory for faces,” Payne observed. “She did not think about the spectra as she classified them—she simply recognized them.” When she needed to concentrate, she would disable her hearing aid.
Leavitt’s feelings about her own work have gone unrecorded. No revealing confessions or letters have been found, just the occasional anecdote. One day, confronted with a particularly mysterious variable called Beta Lyrae, she exclaimed to a colleague, “We shall never understand it until we find a way to send up a net and fetch the thing down!” She yearned, perhaps, to rise above the columns of numbers and really know the stars. Yet even after her discovery of the Cepheid law, she remained assigned to routine photometry, more astronomical needlework. As Cecilia Payne later put it, “Pickering chose his staff to work, not to think.”
Perhaps this would have changed with Shapley. More than anyone, he had seized on Henrietta Leavitt’s stars to plumb the depths of space. He later called her “one of the most important women ever to touch astronomy.” Considering how very few women there were in the field, it is hard to gauge how this weighed on his scales of praise. This was a man who measured the computational difficulty of astronomical jobs in “girlhours” and the really difficult ones in “kilo-girl-hours.”
Any chance the two might have had to collaborate was short-lived. Leavitt, still living with her mother on Linnean Street, was sick again, this time with cancer.
“Took flowers to Miss Leavitt who is very ill,” Cannon wrote in her diary for November 6, 1921. It was a dreary month. By Thanksgiving, Cambridge was besieged by the worst ice storm in memory. Trees and electric poles were breaking under the clinging sleet. The observatory lights went out.
Cannon’s diary describes what came next:
December 6. Went to see poor Henrietta Leavitt, dying with a malignant stomach trouble. So thin & changed. Very, very, sad.
December 8. Clear and cold.
Shapley dropped by to pay his respects. “One of the few decent things I have done was to call on her on her death bed,” he later said. “It made life so much different, friends said, that the director came to see her.” Maybe so.
December 12. Rainy day pouring at night. Henrietta passed away at 10.30 p.m.
December 13. Mr. Leavitt, Henrietta’s brother, called early in morning. Snowy, sloppy, dark day.
December 14. Wednesday. Henrietta’s funeral at Chapel of 1st Cong. Church 2 p.m. Coffin covered with flowers.
She was buried at Cambridge Cemetery (across from the better known Mount Auburn), in the Leavitt family plot. Sitting at the top of a gentle hill, the spot is marked by a tall hexagonal monument, on top of which (cradled on a draped marble pedestal) sits a globe. Her uncle Erasmus and his family are also buried there, along with other Leavitts. A plaque memorializing Henrietta and her two siblings who died so young, Mira and Roswell, is mounted directly below the continent of Australia. Off to one side, and more often visited, are the graves of Henry and William James.
A few days before her death Leavitt had written out her will in longhand, leaving her mother an estate of odds and ends:
Bookcase and books $5
Folding screen $1
Rug $40
Table $5
Chair $2
Desk $5
Table $5
Rug $20
Bureau $10
Bed-stead $15
Mattresses (two) $10
Chairs (two) $2
One @ $100 face value First convertible 4% Liberty Bond $96.33
One @ $50 face value Fourth 4¼% Liberty Bond $48.56
One @ $50 face value Victory 4¾% Note $50.02
The total appraised value came to $314.91.
Also left behind was a photometric survey of the southern sky, and a study of the light curves of novae, including, as a Harvard annual report later put it, “the famous new star of 1918,” which had flared in the constellation Aquila. And she was not quite done with another round of revisions to her magnum opus, the North Polar Sequence. When the International Astronomical Union held its first general assembly in Rome the following May, the Commission of Stellar Photometry, of which she had been a member, recognized her “great service to astronomy.” “She was one of the pioneers in a difficult field of investigation in which she worked with conspicuous success, and it is deeply regretted that she was unable to finish this her last undertaking.”
The next year a Harvard administrative report noted, in passing, something that may have mattered to her more: “She had hardly begun work on her extensive program of photographic measures of variable stars.”
Shapley gave her desk to Cecilia Payne. He tried to persuade her to take over Leavitt’s unfinished projects, but she had other ideas. After completing a celebrated dissertation on the chemical composition of stars, Payne earned Harvard’s first doctorate in astronomy and, under her married name, Cecilia Payne-Gaposchkin, went on to become a full professor and chair of the astronomy department. She never got to meet Leavitt, but she was touched by her story and came to believe that she had been done a great wrong.
“I heard it said when I came to Harvard that what really killed Miss Leavitt was Pickering’s requirement that she devise a method by which the photographic magnitudes determined with all the Harvard instruments could be reduced to the same photometric system,” she wrote years later. “I cannot believe that he made so unrealistic a request.” The judgment seems extreme. This was Leavitt’s North Polar Sequence, which she had taken such pride in. But it did keep her away from her Cepheids.
Payne could understand, from a managerial perspective, that it made sense to assign the best of the assistants to tasks that, however onerous, had to be done. “But it was also a harsh decision,” Payne wrote, “which condemned a brilliant scientist to uncongenial work, and probably set back the study of variable stars for decades.”
Four months after the funeral, Annie Cannon found herself on a steamer bound for Peru, for a tour of the Andes and a visit to the observatory in Arequipa. One evening, after immersing herself in the clear southern skies, she made a note in her diary: “Magellanic Cloud (Great) so bright. It always makes me think of poor Henrietta. How she loved the ‘Clouds.’ ”
3
Like Shapley, Heber Curtis had also moved up in the world, becoming director of the Allegheny Observatory near Pittsburgh. His successor at Lick, a young Swede named Knut Lundmark, carried on the tradition of antagonizing Shapley, claiming that he had been able to pick out individual stars in a pinwheeled nebula called Triangulum or M33 (after the number given it in the eighteenth century by the French astronomer Charles Messier). Assuming (1) that these were the brightest stars in the nebula (which is presumably why he could see them) and (2) that they had the same average magnitude as the brightest stars in the Milky Way, he estimated that M33 was more than a million light-years away. It was, in other words, a full-fledged galaxy.
Shapley quickly challenged him in a letter. Why had his paper not mentioned the work of his friend Adriaan van Maanen, who had measured the rotation of that very spiral, showing it must be small and nearby, or else spinning at impossibly high speeds? Lundmark replied diplomatically and Shapley, for now, was mollified. But he couldn’t resist firing off one of the sarcastic slights that were becoming his trademark: “Whether or not you care to recognize that [van Maanen’s] measures, if real, practically eliminate the ‘island universe’ hypothesis, which you seem to espouse at present time more strongly than any one, is not a matter I can properly concern myself about.”
Lundmark was not so easily defused, as Shapley learned to his annoyance when he picked up an astronomical journal a few months later and read a new paper entitled “On the Motions of Spirals.” The young astronomer was pushing island universes even more vigorously than before, directly questioning the validity of van Maanen’s measurements.
Others had also uncovered problems with the data. The British astronomer James Jeans had studied the van Maanen rotations and found that they violated the known laws of physics. Still he was inclined to believe them (they supported his own theory of how galaxies evolve), explaining away the discrepancies with no less than a proposed modification to Newton’s law of gravity.
If Lundmark was right, van Maanen’s findings were riddled with inconsistencies. He was not suggesting that the astronomer had been sloppy. Measuring such tiny displacements was maddeningly subtle work and open to interpretation. When an object takes 100,000 years to make a single revolution, as van Maanen had concluded, it is not going to move very much in a single year or a decade, or even in the flash of a human lifetime.
No one understood this better than Lundmark. Months later, when he was remeasuring the images of M33, he briefly convinced himself that it really was spinning, so rapidly that “the situation seemed to be rather hopeless for the followers of the island universe theory.” Shapley, of course, was delighted. But the crisis of confidence quickly passed. Closer scrutiny assured Lundmark that he had succumbed to an illusion: there was no sign from this great distance that M33 was rotating at all.
4
Back at Mount Wilson, Hubble was training his sights on Andromeda. It was October 4, 1923. After taking a time exposure of the nebula, he saw what he suspected was another nova. The sky was hazy, so he tried again the next night. This time there seemed to be no question. A second photographic plate showed what appeared to be three novae.
In the observatory office down the mountain in Pasadena, he compared his plate with previous ones taken by Shapley and others. The confirming sign of a nova would be a bright spot appearing where none had been before. One of his flares, however, behaved much differently. Over a period of about a month it had brightened, dimmed, and then brightened again. This was a far more important finding than Hubble had expected. He marked the plate “VAR!” and in February wrote to Shapley: “You will be interested to hear that I have found a Cepheid variable in Andromeda.” According to the period-luminosity scale that Shapley himself had calibrated—Shapley’s curve—the spiral must be a million light-years away.
Hubble went on to boast that, in fact, he had found two Cepheids and nine novae in Andromeda and expected more to come soon. “Altogether next season should be a merry one and will be met with due form and ceremony.”
There he was trying to sound like an Oxford don again. Cecilia Payne later recalled being in Shapley’s office when the dispatch arrived: “Here is the letter that has destroyed my universe,” she remembered his saying, a bit melodramatically. One wonders whether Payne’s memory was a little foggy and the visit actually came later, for Shapley’s immediate reaction was hardly one of defeat.
“Your letter telling of the crop of novae and of the two variable stars in the direction of the Andromeda nebula is the most entertaining piece of literature I have seen in a long time,” he replied a few days later. Continuing in this vein, he argued that Cepheids with periods as long as a month, like the one Hubble had used for his distance measurement, were unreliable as standard candles. It was most likely that what he had found was not a Cepheid at all. False Cepheids, Shapley said, were discovered all the time. He could show Hubble some examples, if he’d like, from the Harvard plate collection. It was going to take a lot more to convince Shapley that his map of the universe was wrong.
Hubble kept on looking at the sky. Pointing the 100-inch telescope toward the constellation Sagittarius, he zoomed in on an irregular-shaped patch of light that resembled a smaller, dimmer version of the Magellanic Clouds. It had first been spotted in the mid–1880s through a 5-inch telescope by Edward Emerson Barnard, an amateur astronomer with so keen an eye that he was given a fellowship at Vanderbilt and later hired as a professor at the University of Chicago. Later observations showed that Barnard’s discovery (usually called by its New General Catalog number, NGC 6822) was a cluster made up of several smaller nebulae and numerous individual stars.
The question, of course, was whether this conglomeration was part of the Milky Way. During 1923 and 1924 Hubble took some fifty pictures of Barnard’s cloud, then compared them with images from earlier years using a blink comparator. As he flipped back and forth between the two plates, variable stars pulsed like traffic lights. He found fifteen of them, concluding that most were Cepheids. According to the period-luminosity scale, what could now almost unequivocally be called Barnard’s Galaxy was 700,000 light-years away.
Hubble also found more Cepheids in Andromeda and in its neighbor M33. This time he broke the news to Shapley with the gentleness of someone who knew he had changed astronomy. Allowing that it was premature to draw final conclusions, he noted that “the straws are all pointing in one direction and it will do no harm to begin considering the various possibilities involved.”
Shapley understood that this was an understatement.“I do not know whether I am sorry or glad,” he replied. “Perhaps both.”
In a later paper Hubble remarked that NGC 6822 indeed appeared to be “a curiously faithful copy” of the Magellanic Clouds. The galaxy had the same general shape and structure. It was just smaller and dimmer. If one trusted the Cepheids to measure the distances, the reason became clear. Barnard’s Galaxy was farther away. For Hubble this consistency was vindication not just of the Cepheid yardstick but of an even grander principle: “The principle of the uniformity of nature thus seems to rule undisturbed in this remote region of space.”
5
With few exceptions, the astronomical world almost immediately recognized that the island universe debate had come to an end. Even Henry Norris Russell realized that he had backed the wrong horse. He urged Hubble to announce his findings at the annual meeting of the American Astronomical Society, to be held jointly in Washington with the American Association for the Advancement of Science. In recent years AAAS, for short, has lost its significance as a place to unveil important new science. The huge annual meetings are devoted primarily to educational sessions, giving scientists and reporters a chance to catch up on developments in various fields. For many scientists and science writers the conference is mostly an opportunity to socialize. But in 1925, the meeting was, as Russell put it in a letter to Hubble, “a splendid forum for a major scientific announcement.” Russell also assured his young colleague that he was a natural for the recently established AAAS Thousand Dollar Prize for the most outstanding paper of the previous year. He was exasperated when, after arriving in Washington for the meeting, no paper from Hubble had been submitted. It arrived, however, at the last moment, and Russell himself read it from the floor. (In the end Hubble shared the prize with the author of two papers on protozoa inside the digestive tracts of termites.)
An abstract of “Cepheids in Spiral Nebulae” was published in May 1925, more than a year after Hubble had first broken the news to Shapley. The reason for the delay, Hubble told colleagues, was the flat contradiction between his results and van Maanen’s. Rechecking his data once again, van Maanen continued to insist that he saw a rotation. But the closer Hubble looked, the more inclined he was to join Lundmark in concluding that the movement was spurious. It was a phenomenon visible to only one man.
To this day no one quite understands where van Maanen went wrong. Perhaps the most compelling theory is that the images of these stellar pinwheels and whirlpools look as though they should be turning (as indeed they are, though many times more slowly). Van Maanen may have been subconsciously influenced by his expectations, seeing what he expected to see.
Why Shapley continued to embrace van Maanen’s theory, until it was no longer possible to do so, requires no elaborate analysis. Cecilia Payne heard him explain it years later: “After all, he was my friend.”
Had Shapley stayed at Mount Wilson, booking time on the mighty 100-inch scope, this new grander universe might have been his own. People would turn on their televisions decades later and admire the glorious photos taken by the orbiting Shapley Space Telescope. Instead he is primarily remembered, a little unfairly, as a great astronomer who couldn’t see beyond the galaxy, who convinced himself that there could be nothing but the Milky Way.
In an interview decades later, he claimed to have forgotten all about the Great Debate. But as he looked back to the 1920s, the details, a bit scrambled, slowly emerged. His most vivid memory was the false one of Einstein’s being there. Shapley said he was surprised that historians were now making so big a deal about the event, contending, a bit disingenuously, that on the “assigned subject matter”—the scale of the universe— he was the clear winner. “I was right and Curtis was wrong on the main point—the scale, the size. It is a big universe, and he viewed it as a small one.”
But that seems in retrospect a minor point compared with what Curtis got right—that our galaxy, no matter how large or small, is one among a multitude, a small outpost in what Hubble would come to call “the realm of the nebulae.” His protégé Allan Sandage later put it like this: “What are galaxies? No one knew before 1900. Very few people knew in 1920. All astronomers knew after 1924.”