The Glass Universe

by Dava Sobel

  Pickering’s reply to Mrs. Draper assured her that all the photographs taken with Draper instruments belonged to the Draper Memorial. “Of course,” he reminded her, “each photograph becomes, like a book, a storehouse of information. It may thus be constantly consulted in future years. This is being done every day with many of the Draper chart photographs, at the expense of other funds. In the same way, in the study of the Draper Memorial of the variables discovered from their spectra, constant use is made of the great number of photographs taken with the Boyden, Bruce, and other instruments.”

  Pickering also stressed the dedication of the Draper Memorial employees. “You will be interested to hear that Mrs. Fleming, not contented with working all day at the Observatory, has undertaken to continue the preparation of the Southern Draper Catalogue in the evenings at her home. A measuring apparatus has been made for her and a recorder provided.”

  Mollified, Mrs. Draper wrote back, “I regret to hear that Mrs. Fleming is taking up night work. I appreciate her zeal and interest but fear she may over tax her strength and break down from over work— I should prefer to hear she intended to take a long vacation.” Mrs. Draper herself was preparing to sail for Europe in July. Before leaving, she thought she should revise her will, so as to guarantee the permanent endurance of Henry’s memorial.

  “My purpose in providing funds for this Memorial has been, as you know, to perpetuate the name of Dr. Draper (my husband) in connection with original work in Astrophysics and especially in the photographic study of stellar spectra, and to contribute to the increase of knowledge in this department of Astronomy.” But now she worried whether, “in the course of years,” this line of work might be exhausted, and another avenue opened.

  “In providing for the continuance of the work in my will, I have to remember that in a relatively short time, you and I will no longer be living, and I have to guard against the possibility that your successor may not be interested in this branch of research, and may prefer to use the fund in some other direction, if he can. In this he may be wise, but I am not disposed to trust his judgment alone, or that of the Trustees of Harvard University, which would be practically the same thing.” She thought a committee of competent astronomers might be appointed to make appropriate decisions when the time came.

  “I appreciate highly the aid which you have given me during the past 17 years,” she reminded Pickering, “and the great amount of time and thought which you have bestowed on the work of this memorial. I believe its results are even now of the greatest interest, and whatever value they have is due to you.”

  • • •

  AT NIGHT IN THE WEST WING, at the eyepiece of the 6-inch telescope, Miss Cannon judged the brightness of the variable stars in her keeping. She used the time-honored technique, fathered by variable star pioneer Friedrich Wilhelm Argelander, of comparing each variable to nearby stars that were either a little bit brighter or just a tad dimmer. The smaller the difference between the target and its neighbor, the better her estimate. It was useless to attempt direct comparison by eye between the ultrabright and the palest light, but the human retina could reliably gauge differences of one-tenth to one-half of a magnitude. Some of Miss Cannon’s stars varied within that narrow range, and could be contrasted with the same comparison star at every stage. For target variables that changed over time by wider margins than a fraction of a magnitude, Miss Cannon recruited two or more neighbors to use as benchmarks. She gave each star a numerical code name tied to its location, and recorded all nuances in the accepted shorthand.

  Miss Cannon was not alone in her solitary pursuit. A few yards away, out on the iron balcony surrounding the dome of the Great Refractor, the junior man, Leon Campbell, tracked his assigned variables through the portable 5-inch telescope, or sometimes a field glass, and often by eye alone. All over New England—all over the country, really, and in foreign countries as well—other variable star observers were engaged in the same occupation. Many amateur astronomers had been enchanted by one of Pickering’s pamphlets, and acted on his suggestions as to which variables they should follow. At least once a month, when their local weather and the phase of the Moon permitted, the members of this volunteer army assessed their stars’ brightness, gauging magnitudes by comparison just as Miss Cannon did, and sent their observations to her at Harvard. She knew some of the more serious contributors by name, such as Frank Evans Seagrave, who owned a private observatory in Providence, Rhode Island, and Mary Watson Whitney, professor of astronomy and director of the student observatory at Vassar.

  Under Harvard’s huge central dome, at the controls of the latest photometer, Oliver Wendell tracked the variables’ slighter fluctuations—as small as three-hundredths of a magnitude. The director stood right beside him. Pickering remained devoted to these observations, and kept a count of the number of stellar assessments he made with each photometer he built. On the night of May 25, 1903, he recorded a personal milestone, his millionth photometric “setting,” in the logbook. Pickering, afflicted with tuberculosis as a youth, had been warned against nighttime exposure at the start of his astronomical career, but by now he could boast he had discovered the fresh-air cure for consumption.

  The stars, Pickering knew, were telegraphing important behavioral clues by their variations in magnitude. Just as the patterns of spectral lines revealed stellar chemical components, so, too, the range of brightness changes over time concealed underlying truths, the nature of which had yet to be grasped. For now, one had merely to track and record the changes, trusting that someday the myriad readings would yield to interpretation. Pickering was never one to speculate when there were data to accumulate.

  Miss Cannon collected all the magnitude determinations made by coworkers and correspondents, and merged them with those of enthusiasts at overseas observatories, from Potsdam to Cape Town, who published their results in professional journals such as the Astronomische Nachrichten and the Monthly Notices of the Royal Astronomical Society. Since 1900, when she took charge of Harvard’s variable star card catalogue, she had added twenty thousand new index cards. In 1903 she turned the whole unwieldy database into a series of tables that could be read by any interested party. Miss Cannon’s opus, “A Provisional Catalogue of Variable Stars,” appeared in the Annals and enjoyed immediate wide distribution.

  Numerous other variable star catalogues, including three by Seth Carlo Chandler, had preceded Miss Cannon’s, and yet she called hers “provisional.” The term bowed to the quickening pace of ongoing discovery made possible by photography. A pre-photography catalogue, published in Vienna in 1865, had listed the 113 variables then known. Miss Cannon’s volume contained 1,227. More than half of these (694) had been discovered on the Harvard glass plates: 509 in Solon Bailey’s globular clusters of the Southern Hemisphere, and 166 by Mrs. Fleming, who spotted their signal bright hydrogen lines as she analyzed stellar spectra for the Henry Draper Memorial.

  Miss Cannon’s tables digested a vast quantity of information, from the position of each variable and its name or other designation to its maximum and minimum brightness, its period, and its spectral category in the Draper classification. One column specified the nature of each star’s variability—whether it was a onetime wonder of the nova type, for example, or a regular repeater of short or long period. Here Miss Cannon relied on the system framed by Pickering in 1880, which divided variables into five varieties.

  Like playing cards in a cosmic game of patience, the stars could be shuffled and dealt in various ways. One might sort them by “suit,” so to speak, according to their spectra, or by the numbered “face value” of their brightness as expressed in magnitudes. The five types of variable stars could be represented by the picture cards—jacks, queens, kings, aces, and wild.

  More than half of the thousand-plus stars in Miss Cannon’s catalogue fell into Pickering’s Type II, the long-period variables. These took up to a year, or even longer, to cycle through their variations. She cou
ld not fathom the causes of their slow fluctuations in magnitude—or of the more rapid rises and dips typical of Types I, III, and IV. Only the relatively rare Type V variables changed brightness for a known reason. These were “eclipsing binaries,” meaning closely orbiting stars that took turns blocking each other’s light. The Type V paradigm, called Algol, or “the ghoul” in the constellation Perseus, dropped from magnitude 2.1 to 3.5 every three days, when the dimmer member of the pair passed in front of the brighter one. The resultant partial eclipse lasted ten hours, after which Algol brightened again, right on schedule. Its regular and pronounced brightness changes, plain to the attentive naked eye, had attracted observers since the 1600s, and earned Algol the common nicknames of “winking star” and “demon star.”

  Mrs. Fleming, who liked to design and sew doll outfits for hospitals and fairs, portrayed Algol’s dual personality as part of a series of astronomical dolls she crafted for the Arequipa families one Christmas, by pairing a large, male Algol figure with a miniature black Dinah doll. In May 1902 she discovered an Algol-type variable herself, while chasing a comet’s path through the glass plates. Her find was the most recent entry in Miss Cannon’s provisional catalogue. Unfortunately, Mrs. Fleming had no suitable spectrum with which to ascertain the new variable’s Draper classification. Although she had graded Algol itself B 8 A, and most of the other twenty-two Algol-type stars A, it was too soon to say where hers fit in the Harvard/Draper scheme. Miss Cannon was content to leave a blank space in the spectrum column alongside the latest Algol discovery. A good number of other blank spaces in her tables pointed up other lacunae, such as missing minimum values, uncertain periods, absent spectra, or questionable variable type. But that was the point of a provisional catalogue, was it not? To expose the gaps in knowledge?

  • • •

  HENRIETTA LEAVITT, THE RADCLIFFE ALUMNA and onetime assistant, came back to Cambridge in the fall of 1903. She had traveled twice to Europe and worked a few years as an art assistant at Beloit College in Wisconsin, near her family’s current home, before realizing how much she missed the observatory. When she wrote to Pickering to say so, he offered her thirty cents an hour to return—a compromise between his regard for her abilities and the standard computer hourly rate of twenty-five cents. On these terms, she joined the new cadre of Carnegie-funded “readers.”

  Despite the earlier bonhomie of Mr. Carnegie, his eponymous institution abruptly ended its support of the Harvard Observatory in December 1903. With scant prospect for the grant’s renewal, Mrs. Fleming had to dismiss the corps of newly trained assistants—all except for Miss Leavitt. Pickering freed other money to pay her salary as a full-time interpreter of texts in the plate library. For her first solo reading assignment, he gave her the Great Nebula in Orion.

  The Orion Nebula, the central gem in the Hunter’s sword, had been meticulously drawn and mapped by George Phillips Bond and famously photographed by Henry Draper. It remained a mysterious thicket of stars embedded in dark lanes of what appeared to be dust and gas. Recently Max Wolf of Heidelberg had studied the nebula and found it studded with variables. Someone needed to follow up Wolf’s observations to confirm the variability of all those stars. Pickering believed he had the right someone in Miss Leavitt, as well as an incomparable collection of long exposures, some lasting several hours, to facilitate her search. Since Orion could be seen from both the Northern and Southern hemispheres, all the Harvard telescopes had photographed it, over a time span of more than ten years.

  Miss Leavitt waded into the nebula, armed with an ingenious aid to brightness estimation—a little glass rectangle containing pictures of model stars at various magnitudes. This diminutive reference guide, about one inch by three, framed in metal and attached to a long handle, resembled nothing so much as a miniature fly swatter. Miss Leavitt liked to call it a fly spanker, since it was “too small to do a fly much damage.” Within six months, she had confirmed sixteen of Wolf’s variables and found more than fifty new ones, which were confirmed in turn by Mrs. Fleming.

  Miss Leavitt achieved her next spate of discoveries by a different method. Edward King, the observatory’s master photographer, made her a single positive plate from one of the many glass negatives of the Orion Nebula. On the glass positive, the stars shone white against a grainy gray background. Miss Leavitt superposed each negative on this positive, then examined the combination through a magnifying loupe. Unchanging stars tended to neutralize each other, but eight new variables popped out at her. Two months of this effort added another seventy-seven variables to Miss Leavitt’s life list. She moved on to other nebulae in or near other constellations. She amassed another two hundred variable stars in the two nebulae seen by Ferdinand Magellan when he circumnavigated the globe in the 1520s. To Magellan’s eye, they had resembled a pair of luminous clouds afloat in the southern night sky. Astronomers who later resolved the clouds into star clusters still called them by Magellan’s name. In early 1905, in the Small Magellanic Cloud alone, Miss Leavitt uncovered nine hundred new variables.

  “What a variable-star ‘fiend’ Miss Leavitt is,” Charles Young of Princeton wrote in awe to Pickering on March 1, 1905. “One can’t keep up with the roll of the new discoveries.” Mrs. Draper expressed similar sentiments about “Miss Leavitt’s remarkable discovery of variables,” on March 11. As the tally continued to rise, Mrs. Draper wrote again in May to applaud “the large number of variables in the Small Magellan Cloud. It is certainly strange that so many of them should be found apparently close together. Will you please congratulate Miss Leavitt for me.” She offered congratulations also to Pickering’s brother, William, “upon the discovery of the tenth satellite of Saturn, [as] he is now proprietor of two of the planet’s attendants.”

  William had waited four years for other astronomers to second his discovery of Phoebe, the ninth satellite of Saturn. By 1904 the little moon had been seen through several large telescopes other than the Bruce, and also proven to ply the most unusual orbit in the solar system. Phoebe circled Saturn in a backward, or retrograde, direction, against the traffic flow of the other moons and rings. This discovery led William to the inescapable conclusion that Phoebe had started life as an errant asteroid. When she wandered too close to Saturn, the giant planet had captured her and constrained her in a retrograde orbit.

  William’s success with Phoebe spurred him to examine more Bruce plates of Saturn’s surroundings for signs of additional satellites. When he found what he believed to be the tenth moon, on April 28, 1905, he named it Themis, another Titaness of Greek mythology. He struggled with the calculation of its orbit, but none of the computers could be spared to assist him. As usual at the Harvard Observatory, the stars took precedence over the planets.

  Edward Pickering kept fellow astronomers abreast of Miss Leavitt’s advances by issuing a rapid-fire series of circulars. A few of these publications included small prints showing parts of the images she inspected. The prints, magnified for clarity and crowded with tens to hundreds of thousands of stars, conveyed the flavor of her Herculean task, which included approximating the percentage of variables per plate. “It is very difficult to count the faint stars which cloud the background,” Pickering said with obvious understatement, “on account of their closeness to one another, and the number is certainly underestimated.”

  Occasionally Pickering described one of Miss Leavitt’s finds as a nova or an Algol variable, but the great majority of her variables demonstrated only slight changes of about half a magnitude over very short time scales. Perpetually aflutter, they cycled from maximum to minimum at least once every day. Their rapid fluctuations engendered a new photographic approach, involving several successive short exposures on the same plate, so that each star appeared as a series of dots. When Mrs. Draper visited for a couple of days, she professed herself more impressed than ever with the “vast work” accomplished. “You live so constantly in the midst of it,” she wrote to Pickering on May 29, 1905, “that you can not
so well appreciate it.”

  • • •

  PICKERING REPRISED HIS LIBRARY METAPHOR in the fall of 1905, to deplore the underutilization of the plate stacks. The nearly two hundred thousand glass “volumes” drew only twenty readers to their contents, and the director was desperate for more. Harvard president Charles Eliot agreed to help. “I expect to be in Mr. Carnegie’s house in New York on the 15th of November,” he wrote Pickering from his office in the Yard, “and will then promote a visit by Mr. and Mrs. Carnegie to our Observatory, if I have the opportunity.”

  Mrs. Fleming also tried to revive the philanthropist’s interest in the plate library. She sent a long letter to his wife, Louise Whitfield Carnegie, along with a small gift.

  Mrs. Carnegie responded warmly from the family’s vacation cottage in Fernandina, Florida, on January 11, 1906: “We left New York three days after Christmas and have been so occupied getting ourselves established here for the winter that I am only just able to turn my thoughts to the many kindnesses showered upon us at the Christmas season by our friends. Chief among these kindnesses is your own great kindness to us—The Wonderful ‘Story of the Stars’ in its dainty Christmas dress, with the remarkable lantern slides, form one of the most unique lovely greetings that can be imagined, & when we think of your kindness in taking all this thought for us, words fail me to express how deeply we are touched & how truly we appreciate this precious gift. Then too, may I add, when this gift is the original work of the great discoverer herself, are we not indeed a favored people and are we not blessed in knowing her? We are so proud of you as a Scotchwoman! & proud of your womanhood! For I believe the feminine mind is more capable of ‘fathoming the Eternal Thought’ just as the feminine heart nestles closer to Nature & to Nature’s God.”