The Glass Universe: The Hidden History of the Women Who Took the Measure of the Stars

Henry had always preferred reflecting telescopes, which gathered light by means of a mirror in lieu of a lens, over refracting ones that could introduce spurious color effects. He had begun crafting his own mirrors right after medical school, and must have made a hundred in all, but the 28-inch was his great reflector. On November 12, 1867, the day after he and Anna exchanged marriage vows in her father’s living room, they went downtown together to shop for a glass disk—the kind used in skylights—large enough to form a mirror 28 inches across. They referred ever after to that excursion as “our wedding trip.” It took them years to grind and polish the disk to the desired curvature and apply the ultrathin coat of silver that transformed the glass into a perfect mirror.

The 28-inch reflector had enabled them to take their landmark first picture of the spectrum of Vega in 1872, as well as their unrivaled photograph of the so-called Great Nebula in Orion ten years later, and also their final series of stellar spectra images during the summer before Henry’s death. On one of those humid July nights, undone by overcast skies, the two of them had quit the observatory around midnight to retire. But as they neared their country house two miles away on Wickers Creek in Dobbs Ferry, they saw the clouds dissipating, so they turned the horses around and drove back to Hastings to resume their work. She remembered returning that way on numerous other occasions just to seize a few more hours—even long ago, when they thought they had all the time in the world.

• • •

“MRS. DRAPER HAS DECIDED to send to Cambridge a 28-inch reflector and its mounting,” Pickering announced on March 1, 1887, in the first annual report of the Henry Draper Memorial. He praised the project’s benefactress for providing not only the instruments required for the project but also the means for keeping them actively employed by operators “during the whole of every clear night,” and for “reducing the results by a considerable force of computers,” and for publishing them as well. He hoped that other donors would follow her example by similarly endowing astronomy departments elsewhere with the means to function to their fullest.

In the spring of 1887, while Mrs. Draper negotiated with the Hudson River Railroad for a car to carry the 28-inch to Harvard, the observatory received another huge bounty—approximately $20,000, to be augmented by $11,000 annually—for the establishment of an auxiliary station on a mountaintop.

Pickering had been climbing mountains all his life. He began summiting in New England with youthful companions who called him “Pick” and even “Picky.” He later measured the heights of points of interest in New Hampshire’s White Mountains on solo treks with fifteen pounds of apparatus strapped to his back. In 1876, around the time he left the MIT physics department to direct Harvard’s observatory, he founded the Appalachian Mountain Club for fellow outdoorsmen, and served as its first president. Still an active member in 1887, he could well imagine the advantage of stationing a telescope at high altitude.

The source of the sudden windfall was the contested will of Uriah Boyden, an eccentric inventor and engineer who had received an honorary Harvard degree in 1853. When Boyden died in 1879, unmarried and childless, he allotted $230,000 to perch an observatory far above the atmospheric disturbances that plagued astronomers at sea level. Many noble institutions, including the National Academy of Sciences, vied for control of the Boyden estate, but Pickering convinced Boyden’s trustees that Harvard University was the most likely of the suitors to invest the money wisely, and the Harvard Observatory most fit to carry out the testator’s instructions. Triumphant after five years of polite wrangling, Pickering organized an exploratory expedition to the Colorado Rocky Mountains.

The Boyden Fund gave Pickering the means to hire his younger brother away from MIT. William, likewise a charter member of the Appalachian Mountain Club, thus became the director’s assistant and guide for the western reconnaissance. The brothers left Cambridge in June 1887 along with Lizzie Pickering, three junior volunteers from the observatory, and fourteen crates of equipment. Mrs. Draper joined them at Colorado Springs in July.

Although no high-altitude astronomical observatory yet existed in the United States, the federal reservation at Pikes Peak was home to the world’s highest meteorology station, maintained at 14,000 feet by the U.S. Army Signal Corps. This made Pikes Peak the only American mountain where particulars of weather (beyond the statistic of annual rainfall) were known. When Pickering’s party of five men ascended in August, leading mules laden with scientific instruments, they encountered a snow squall, a hailstorm, and a thunderstorm they described as violent. Over the course of the month, they camped and compared conditions on three peaks in the region by various means, such as a sunshine recorder William had modified as a complement to a rain gauge, and also by photographing the sky through a 12-inch telescope. Conditions did not seem optimal. What was worse, rumor had it that Pikes Peak might be turned into a state tourist attraction, and be overrun with non-astronomers.

Pickering returned to Cambridge without having settled the placement of the Boyden Station. He thought he might revisit the Rockies the following summer, or try a different mountain range.

In October, after Mrs. Draper returned East, closed her Dobbs Ferry house for the season, and reestablished herself on Madison Avenue, she thanked Pickering for the summer’s adventure with the gift of an ornamental pocket telescope that had once belonged to King Ludwig of Bavaria.

• • •

WITH TWO AND OFTEN THREE TELESCOPES taking pictures through the night, the observatory devoured plates at a rapid rate. Between 1886 and 1887, advances in the quality of manufactured dry plates extended their recording range to fainter stellar magnitudes, and Pickering took full advantage of each new development. He tried different companies’ wares and shifted suppliers accordingly; he encouraged manufacturers to keep improving the sensitivity of their plates—and to send him their latest products for testing.

The volume of data to be calculated rose in proportion to the number of photographs taken. Anna Winlock’s younger sister, Louisa, assumed her place in the computing room in 1886, and was joined the following year by Misses Annie Masters, Jennie Rugg, Nellie Storin, and Louisa Wells. The staff of female computers now numbered fourteen, including Mrs. Fleming, who served as their supervisor. Most of the ladies were younger than she, more or less her social equals, and respectful of her authority. That situation shifted in 1888 with the addition of twenty-two-year-old Antonia Maury, who was not only a Vassar College graduate with honors in physics, astronomy, and philosophy, but also the niece of Henry Draper.

“The girl has unusual ability in a scientific direction,” Mrs. Draper told Pickering on March 11, 1888, “and is anxious to teach chemistry or physics—and is studying with that object in view.”

As a child, Antonia Maury was allowed into her Uncle Henry’s chemistry laboratory at the big house in New York City, where she “assisted” him by handing him specific test tubes he requested for his experiments. Before she turned ten, her father, the Reverend Doctor Mytton Maury, an itinerant Episcopal minister, taught her to read Virgil in the original Latin. Her mother, Henry Draper’s sister Virginia, was a naturalist enamored of every bird, flower, shrub, and tree on the Hastings property; she had died in 1885 while Antonia was studying at Vassar.

Pickering felt uncomfortable offering the standard computer pay of twenty-five cents per hour to a person of Miss Maury’s achievements. He expressed something like relief when she failed to answer his letter, but Mrs. Draper interceded for her through April and May.

“The girl has been very busy,” the aunt explained. Although Reverend Maury had relocated to Waltham, Massachusetts, for his work, he had neither found a home for his family nor enrolled his two younger children, Draper and Carlotta, in school, leaving Antonia to take charge of these matters. By mid-June she had joined the Harvard corps.

Pickering assigned Miss Maury the spectral measurement of the brightest stars. Mrs. Fleming had worked from plates containing hundreds of spectra crowded together, and on which the bright stars appeared overexposed. The 11-inch Draper telescope focused on just one star at a time. Each spectrum imaged in this manner spread over an expanse of at least four inches, even before enlarging. The gratifying increase in detail gave Miss Maury much to ponder as she examined the plates under a microscope. In the same blue-violet region of Vega’s spectrum where her uncle had photographed four lines in 1879—and ten in 1882—she now counted more than one hundred.

Along with measuring the distances between the lines and converting them to wavelengths, she was expected to classify each spectrum according to Mrs. Fleming’s criteria. But Miss Maury had so much more detail to work with that she could not confine her impressions to those parameters. Some of the lines she looked at were not simply thick or intense, but also hazy or fluted or otherwise noteworthy. Such nuances surely deserved attention, for they might illustrate as yet unsuspected conditions in the stars.

• • •

WHEN HARVARD’S SECOND MOUNTAIN reconnaissance headed West in November 1888, Pickering opted out. He could not possibly afford enough time away from the observatory to fulfill the mission’s ambitious itinerary, which was to begin site testing near Pasadena, California, and continue among the Andes in Chile and Peru. He put his brother, William, in charge. While in California, the team would also visit the Sacramento Valley to observe and photograph the total solar eclipse of January 1, 1889.

Ordinarily, Pickering did not support eclipse expeditions, on practical grounds. He deemed the expense too high, given the high risk of failure. An ill-placed cloud during the scant moments of totality could scotch the whole enterprise (as he had learned firsthand when he went to Spain with former director Winlock for the eclipse of December 22, 1870). But if, as in the present case, the path of totality nearly crossed the path of exploration for the new Boyden Station, Pickering would not object to a small detour.

Favorable weather smiled on the observers for the New Year’s Day eclipse. Excitement at the rare sight, however, shook the astronomers and the large crowd of onlookers alike. At the start of totality, the spectators started to yell. The noise drowned out William’s call to the person counting out the seconds, and his struggle to make himself heard caused him to take fewer pictures than he intended. He also forgot to remove the lens cap from the spectroscope.

From his disappointment in Sacramento, William went south to Mount Wilson, where he and a few assistants were to test atmospheric conditions by observing for several months with a 13-inch telescope they brought along for that purpose. At the same time, the other half of the team departed for South America. In Pickering’s grand scheme, two mountain observatories were better than one. A California aerie would improve on the work done at Cambridge, while an additional satellite station in the Southern Hemisphere would widen Harvard’s field of view to encompass the entire sky.

Pickering entrusted control of the South America venture to Solon I. Bailey, age thirty-four, who had joined the observatory staff as an unpaid assistant two years earlier and quickly proven himself deserving of a salary. Like Pickering, Bailey had a younger brother with a talent for photography, and so, with Pickering’s blessing, Solon appointed Marshall Bailey as his second-in-command, and planned to meet him in Panama after the eclipse. Facing a trip expected to last two full years, Solon took along his wife, Ruth, and their three-year-old son, Irving.

The February 1889 voyage aboard the San Jose of the Pacific Mail gave Bailey occasion to practice his Spanish with several fellow passengers, whose names he recorded in his journal. On deck, he enjoyed watching Venus sink into the sea after sunset, “plainly seen till she touched the water.” In the predawn February sky, he sighted the Southern Cross for the first time. Bailey had loved the stars since his boyhood in New Hampshire, where he witnessed the great natural fireworks of the 1866 Leonid meteor shower. Now he would meet a sky’s worth of new constellations, which prospect inured him to whatever hardships lay ahead.

The bulk of the Andes expedition supplies—everything from photographic plates to prefabricated buildings—traveled with Marshall from New York to the Isthmus of Panama, then overland, past the recently aborted French canal effort and the graveyards of fever victims to another ship bound for Callao, near Lima.

The party rode the Oroya Railroad twenty miles east from Lima to Chosica, and from there the Bailey brothers ascended on foot and by mule to elevations of 10,000 feet or more. Their native guides nursed them through bouts of altitude sickness with an effective local remedy, namely the odor of bruised garlic. No particular peak impressed Bailey as ideal, but he needed to seize the good weather of the dry season, and so settled for a nameless mountain with the least obstructed view. It stood just over 6,500 feet high, barely accessible by a path that switchbacked up and around for eight miles. The Baileys labored alongside a dozen locals for three weeks to improve the route from the hotel in Chosica to the site, and then helped drag eighty loads of equipment up that road to the makeshift observatory. When the family moved in on May 8 along with their Peruvian assistant, two servants, cats, dogs, goats, and poultry, their only neighbors were centipedes, fleas, scorpions, and the occasional condor. They relied on a muleteer for daily supplies of water and food.

The Baileys assessed the brightness of the southern stars with the same meridian photometer that Pickering had used in Cambridge, in order to make their observations exactly comparable to his. Similarly, they photographed the southern stellar spectra for the Henry Draper Memorial with the selfsame 8-inch-aperture Bache telescope that had seen nightly duty through the project’s first two years. Mrs. Draper replaced the original workhorse at Harvard with another of the same specifications.

Solon Bailey stayed in touch with Pickering as regularly as the mails allowed. When he shipped the first two cases of glass plates to Cambridge, he said they came from an as yet unnamed place that he would like to call Mount Pickering.

“Mt. Pickering might wait,” the director wrote back on August 4, 1889, “until I have done as good work as you have on a Peruvian mountain.” With local approval, the Baileys christened the site Mount Harvard instead.

When the October onset of the rainy season halted work on Mount Harvard, Bailey moved his wife and son to Lima, then set off with his brother to scout better locations for a permanent base. It took them four months to find a place that met their requirements, on the high desert plain near the town of Arequipa. At 8,000 feet, the air was clear, dry, and steady, and the nearby volcano, El Misti, was nearly extinct.

• • •

WHILE THE BAILEYS EXPLORED PERU, Edward Pickering became engrossed with the odd spectrum of a star called Mizar in the handle of the Big Dipper. The star had first drawn his surprised attention on a Draper Memorial photograph taken March 29, 1887, which showed an unprecedented doubling of the spectrum’s K line. (Although Fraunhofer’s original lettering ended at I, later researchers added other labels.) Soon after Pickering shared the unusual news with Mrs. Draper, the strange effect vanished as suddenly as it had appeared. Subsequent images of Mizar’s spectrum failed to recover the double K line, but still Pickering kept watching for its return. On January 7, 1889, Miss Maury saw it, too. Pickering, who rarely invoked an exclamation point, wrote Mrs. Draper, “Now it seems nearly certain that it is sometimes double and sometimes single!” Although, he quickly added, “It is hard to say what this means.” He suspected that Mizar, also known as Zeta Ursae Majoris, might turn out to be two stars with virtually identical spectra, too closely aligned to be seen separately, even through a big telescope.

Miss Maury could picture the Mizar pair as two wary combatants, circling each other while vying for advantage. Her distant vantage point made it difficult to distinguish the two separate bodies—impossible, in fact, when either one stood in front of the other along her line of sight. But Mizar’s twin fighters were emitting light. As they revolved, their relative motions slightly altered the light’s frequency: the approaching starlight shifted slightly toward the blue end of the spectrum, the receding starlight toward the red. Those shifts added up to the small K-line separation that created the doubling effect.

Pickering and Miss Maury tracked Mizar’s K line through months of ambiguous changes, until they saw the doubled line again on May 17, 1889. Photographs taken a few nights before and after the doubling portrayed the line as hazy—somewhere between single and double. Miss Maury had been wise to trust her intuition about hazy lines.

That Sunday, on her day off, Miss Maury wrote to her aunt, Ann Ludlow Draper, the wife of Henry’s brother Daniel. Everything she reported in her long, chatty letter seemed to touch on the theme of single and double. On a visit to the Boston Public Garden she had seen “a wonderful display of tulips single and double of all colors.” She now had dual Vassar Alumnae Association membership in both the Boston and New York branches. “I told them I should have a chance to vote twice but they didn’t seem to be afraid.” She saved the most interesting case for last:

“Tell Uncle Dan that the other day Prof. Pickering succeeded in photographing the double K line of Zeta Ursae Majoris. Other lines were also double that at times are single so I suppose his theory is proved that the change is due to the rotation of two close stars of the same type around one another. It is a very pretty thing. They have been trying for months to catch it double. Prof. Pickering thinks its period must be about fifty days but has not finished the calculations yet. Of course nothing ought to be said about it publicly till it is all worked out.” She signed the letter “With love, Antonia.”

Pickering wrote a report of the preliminary results, making sure to credit “Miss A. C. Maury, a niece of Dr. Draper” for her careful study of Mizar’s spectrum. He sent the paper to Mrs. Draper, who carried it to Philadelphia for the annual meeting of the National Academy of Sciences, where their mutual friend George Barker read it aloud to the assembly on November 13, 1889. Barker assured Pickering that the K-line news “awakened a lively interest.”

A few weeks later, on December 8, with Mrs. Draper present at the observatory, Mizar’s K line doubled again, right on schedule. Within days, Miss Maury found the double K line in another star, Beta Aurigae (the second brightest in the constellation of the Charioteer). Now there were two examples of newfound star pairs that had been discovered by their spectral characteristics alone. And before the week was out Mrs. Fleming identified a third suspected “spectroscopic binary” on several plates from Peru.

“Now if all these results ensue in consequence of your recent visit here,” Pickering cajoled Mrs. Draper, “is it not a sufficient argument in favor of your coming oftener?”

Mrs. Draper wished she might flatter herself, she replied, “that the interesting results obtained during my visit were in consequence of my being with you; my friends have often called me a ‘Mascotte’ but I fear my luck will not extend so far.” Nevertheless she declared herself “delighted” with the new finds. Additional examples would help convince certain members of the Academy, present at the recent meeting, who “thought our imagination had run away with us.” More confirmation came in an independent discovery of another spectroscopic binary, also in late 1889, by Hermann Carl Vogel of the Potsdam Observatory.

Vogel had been using spectroscopy to answer a different question—not What are stars made of? or How can stars be divided into groups? but How fast do they move toward or away from Earth in the line of sight? By the degree to which certain lines in their spectra shifted toward blue or red, Vogel calculated their radial velocity. Some traveled as fast as thirty miles per second, or well over one hundred thousand miles an hour.

As Miss Maury continued to chart the spectral changes of Mizar, she concluded that its component stars orbited their common center of gravity once every fifty-two days. She deduced an even shorter period of only four days for Beta Aurigae, the spectroscopic binary that she had discovered. Indeed, she could watch the Beta Aurigae spectrum change from one photograph to the next over the course of a single night. She calculated the orbital speeds in the two binary systems. “A mile a minute” sounded rapid to her ear, but these stars were racing around at more than a hundred miles a second. Her uncle Henry had looked to the spectra to uncover the stars’ chemistry, and now the spectra were also yielding the stars’ celerity.

• • •

THE YEAR 1890 SAW THE PUBLICATION of Mrs. Fleming’s opus, “The Draper Catalogue of Stellar Spectra,” in volume 27 of the observatory’s Annals. Pickering rewarded her with a raise in salary and full acknowledgment in his introductory remarks: “The reduction of the plates was begun by Miss N. A. Farrar, but the greater portion of this work, the measurement and classification of all the spectra, and the preparation of the Catalogue for publication, has been in charge of Mrs. M. Fleming.” She styled herself “Mina Fleming” now. In addition to the dedication she had shown in measuring and classifying the spectra of ten thousand stars, she had also expertly proofread the catalogue’s four hundred pages. Most of the pages consisted of tables, twenty columns wide and fifty lines long, representing approximately one million digits in all.

The Draper Catalogue sorted the stars by the appearance of their spectral lines—not merely for the sake of sorting, but in the hope of opening new avenues of investigation. The classification inspired Pickering, for one, to analyze the distribution of stars by spectral type. Peering into the luminous band of the Milky Way, he found a preponderance of B stars. The B stars clustered along the Milky Way as though they had an affinity for one another or for that region of space. The Sun, a G star, seemed to Pickering to have little relation to the lights of the Milky Way.

Meanwhile Miss Maury proceeded with her own elaborate classification system. She intended to increase Mrs. Fleming’s fifteen classes to twenty-two, and also subdivide each type into three or four subcategories, based on the further gradations she detected in the spectra of her bright stars. The strain on her vision prompted her to consult a Boston oculist, who prescribed eyeglasses.

“Dear Auntie,” she wrote to her great-aunt Dorothy Catherine Draper on February 18, 1890, “I am now writing up the results of my work of the last two years. I have made a short outline that is the beginning of my classification. I was very much afraid Prof. Pickering would not like it, but I am glad to find that he is quite satisfied and says with a few changes it will do to print. Of course it will take me a long time to get the whole thing written and I expect all the details will make quite a volume. … I wear your black hat every day and your afghan keeps me warm at night.”

In his fourth annual report of the Henry Draper Memorial, published shortly after Mrs. Fleming’s catalogue in 1890, Pickering announced that the total number of photographs taken with the various telescopes had reached 7,883. Other observatories, he noted, made the “very common mistake” of accumulating photographs without deriving results from them through discussion and measurement. At Harvard, however, a corps of computers had been studying the photographs for several years, so that “for many purposes the photographs take the place of the stars themselves, and discoveries are verified and errors corrected by daylight with a magnifying-glass instead of at night with a telescope.” Here, too, as in the Annals, he cited both Mrs. Fleming and Miss Maury by name. It was the niece of Henry Draper, he emphasized, who had discovered the doubling of the lines in Beta Aurigae.

In line with his usual practice, Pickering distributed the fourth annual report of the Henry Draper Memorial far and wide, including publication in Nature and other scientific journals. The report found one of its most appreciative audiences in England, at the home of astronomer and military engineer Colonel John Herschel. As a grandson of William Herschel (discoverer of the planet Uranus) and a son of Sir John Herschel (thrice president of the Royal Astronomical Society), the colonel had seen his share of important leaps in celestial knowledge.

“I have just rec’d your last H. D. Mem. report,” he wrote to Pickering on May 28, 1890. “It is very like a pudding all plums—but I will ask you to convey to Miss Maury my congratulations on having connected her name with one of the most notable advances in physical astronomy ever made.”

Like the colonel’s much celebrated great-aunt, Caroline Herschel, Miss Maury had entered a field of discovery dominated by men, yet she stood among the first astronomers to detect an entirely new group of objects through the upstart method of spectral photography. Its future—and hers—seemed full of promise.