by Dava Sobel
• • •
MISS CANNON FELT TOUCHED “very vitally” by the adoption issue. “It is a far cry from our own sun to the nearest star,” she wrote in a memoir, “yet we know that the stars are suns and that many of them are in exactly the same state of composition as our sun. It is therefore befitting that the Solar Union should be interested in the composition of the heavenly bodies.” Though she welcomed their interest, she worried “lest this great international body might adopt one of the several other systems of classification proposed, and not adopt ours.”
Mrs. Fleming had founded the first Draper catalogue on thousands of tiny spectra photographed through a prism at the telescope’s objective end. Those pictures portrayed the violet end of the spectrum well enough, but captured very little of the red. Now that newer photographic techniques and improved dry plates could cover a wider spectral range, Miss Cannon tested the soundness and durability of the Draper classification by reexamining some of the old stars in new photographs. She took pains to work “blind,” first classifying the wide new spectra, and only later looking up Mrs. Fleming’s designations. It soothed and gratified her to see the overall agreement between them. Apparently the violet end of the spectrum sufficed to settle a star’s identity. Miss Cannon corrected some of the original classifications, but more often she just enhanced them with the added spectral detail at her disposal, such as changing an F star to an F 5 G.
Mrs. Fleming helped revise the ever-improving Draper Catalogue by revisiting the numerous spectra formerly lumped in the “peculiar” category. The pace of her variable star discoveries remained slow as she hastened the delivery of Annals volumes to the printer. That winter she found “only” eight. Early in 1911, however, in recognition of her cumulative record, the Sociedad Astronómica de México awarded her its Guadalupe Almendaro gold medal for her prowess among the variables. The Bruce Medal still eluded her, but she hardly lacked for recognition from her fellow members of the Astronomical and Astrophysical Society of America, or the fans who had made her an honorary member of the Royal Astronomical Society and the Société Astronomique de France.
Mrs. Fleming so often visited Professor Sarah Whiting’s classes at Wellesley as a guest lecturer that the college had named her its Honorary Fellow in Astronomy. She was anticipating giving another talk at Wellesley, scheduled for late May, when the fatigue that had dogged her all spring turned to malaise. She elected to enter a hospital for rest, but, once there, her condition worsened, and she developed a fatal pneumonia. Edward Fleming, now the chief metallurgist for a large copper company in Chile, could not get to Boston in time to see his mother before she died on May 21, 1911. She was fifty-four years old and had devoted thirty of her years to the observatory.
“I have just this moment, to my very deep regret, seen in ‘Science’ a notice of Mrs. Fleming’s death,” Henry Norris Russell wrote to Pickering from Princeton on June 2. “I cannot do less than write at once to express my sympathy for the loss that I know will be felt very deeply and in many ways by all the circle at the Harvard Observatory, and by a far wider one of friends outside.” The youthful Russell had spent time with Mrs. Fleming at meetings in Cambridge and on the previous summer’s trek to Pasadena for the Solar Union. “Her loss will be a severe one to science, and must be a terrible blow to her friends,” he commiserated. “I did not know the son of whom she spoke so often, and so can hardly send a message to him; but the sense of loss on hearing that she is gone was so keen that it seemed natural to write to you.”
Pickering’s eulogy for Mrs. Fleming in the Harvard Graduates’ Magazine retold parts of the saga she had shared with him over the years about her ancestors, the Claverhouse “fighting Grahams”—how her great-grandmother had eloped with Captain Walker of the 79th Highlanders, followed him to Spain in the Peninsular War, then given birth to a son on the Corunna battlefield the very day the captain was killed in combat. Surely the family mettle had stiffened Mrs. Fleming’s spine. As her longtime supervisor, Pickering could aver, “It was only necessary to tell her exactly what was needed, and she saw that it was carried through successfully in every detail.” After enumerating her many astronomical discoveries and distinctions, he said she “formed a striking example of a woman who attained success in the higher paths of science without in any way losing the gifts and charm so characteristic of her sex.”
Miss Cannon had written the obituary notice for Mrs. Fleming that Henry Norris Russell read in Science, as well as another, longer one for the Astrophysical Journal. The articles gave her occasion to praise her late friend’s “naturally clear and brilliant mind,” “her extremely magnetic personality,” and “that quality of human sympathy which is sometimes lacking among women engaged in scientific pursuits.” Miss Cannon also took pains to describe the rare collection of glass plates entrusted to Mrs. Fleming’s reliable care: “Each photographic plate may be likened to the only existing copy of a valuable book, and, being very fragile, must be safely stored, and at the same time must be accessible, so as to be consulted readily at any moment.”
It seemed right that Miss Cannon should succeed Mrs. Fleming as the official curator of astronomical photographs. Pickering raised the idea in October 1911 with Harvard’s new president, Abbott Lawrence Lowell (brother of Percival Lowell), who took office after Charles Eliot retired in 1909. Not only had Miss Cannon fulfilled the duties of curator since Mrs. Fleming’s passing, Pickering said, but she had done so “in a very satisfactory manner.” Moreover, he added by way of endorsement, “Miss Cannon is the leading authority on the classification of stellar spectra, and perhaps on variable stars.”
Lowell reacted negatively. “I always felt that Mrs. Fleming’s position was somewhat anomalous,” he replied on October 11, “and that it would be better not to make a regular practice of treating her successors in the same way.” He therefore declined to recommend that the Harvard Corporation appoint Miss Cannon. Instead, he suggested, Pickering should install her himself as a matter of ordinary department business, with less fuss, lower pay, and no cause to cite her name in the university catalogue.
The members of the Visiting Committee were appalled. “It is an anomaly,” their 1911 report said of the slight to Miss Cannon, “that, though she is recognized the world over as the greatest living expert in this line of work, and her services to the Observatory are so important, yet she holds no official position in the University.”
Miss Cannon did not let the denial of university title stand in the way of duty. In October 1911 she embarked on new projects to unify and strengthen the Draper system. She reclassified Miss Maury’s bright northern stars, to conform their Roman numerals to the current Draper designations. She took up Mrs. Fleming’s unfinished last catalogue of faint southern stars and brought its 1,688 listings into alignment. Both her speed and her certainty in judgment had increased, as, too, had her love for the work. She thought she might as well keep going, examine more plates, continue classifying indefinitely, expand the Draper Catalogue exponentially.
• • •
PICKERING’S VOLUNTEER ARMY of variable star observers blanketed the Northeast by 1911, and extended as far west as California. There was even an outlier in Australia. Faculty and students at New England colleges such as Amherst, Vassar, and Mount Holyoke participated energetically in the routine observations. Strong foreign support arrived monthly from amateurs in the Variable Star Section of the British Astronomical Association. Harvard’s own professional staff still led the charge, with Leon Campbell alone averaging a thousand observations per month through a 24-inch reflector.
Campbell’s attentions shifted in the spring of 1911, when Pickering sent him to Arequipa as director of the Boyden Station. The new post positioned Campbell to keep vigil over the long-neglected long-period variables of the southern sky, but also forced his abandonment of the northern ones. To fill Campbell’s vacancy, Pickering called on the corps of volunteers. He drew up a list of 374 variables requir
ing frequent surveillance, and assigned each star to one or more regular observers. He also circulated the list as an invitation for others to participate. Given the interruptions to be expected from inclement weather, moonlight, and personal engagements, one star could never have too many pairs of eyes on it. He prepared printed forms to facilitate the filing of reports, provided finder charts to help new recruits locate their stars, and promised to publish the volunteers’ observations. Hoping to head off any needless duplication of effort, Pickering urged his troops to communicate among themselves and cooperate wherever possible, such as by observing at different times of the month and different hours of the night.
Popular Astronomy editor Herbert C. Wilson saw the need for an even higher order of organization among variable stargazers. In the August–September 1911 issue of the magazine, Wilson entreated his readers, “Can we not have in America an association of observers with a ‘Variable Star Section,’ a ‘Jupiter Section,’ etc.?” In almost instant reply, lawyer and avid amateur observer William Tyler Olcott of Norwich, Connecticut, announced the October formation of the American Association of Variable Star Observers (AAVSO).
Olcott had caught the variable star fever from Pickering, at a public lecture the director gave in 1909. The two corresponded afterward, and Pickering, recognizing Olcott’s dedication, arranged for Leon Campbell to coach him at his Connecticut home. The founding of the AAVSO cemented the already close ties between Olcott and Harvard.
Professor Anne Sewell Young of Mount Holyoke, one of Pickering’s most reliable regulars, immediately signed up as a charter member of Olcott’s association. In December 1911 her recent observations formed part of the AAVSO’s first published report in the pages of Popular Astronomy. Soon Sarah Frances Whiting and her assistant Leah Allen of the Wellesley College observatory joined the AAVSO, and also Maria Mitchell’s successor at Vassar, Caroline Furness. The group welcomed devotees from any sort of day job. Charles Y. McAteer, for example, worked as a locomotive engineer for the Pittsburgh, Cincinnati, Chicago and St. Louis Railway Company. At the end of the night freight run into Pittsburgh, he would go home to the 3-inch telescope in his backyard and observe variables till dawn.
The AAVSO members concentrated on the variables of long period. Most such stars waxed and waned gradually through as many as nine magnitudes of change in the course of a few months to a year-plus. They were always on their way up or down the brightness scale, filling hours with quiet purpose for Pickering’s minions. The short-period variables, on the other hand, defied tracking by telescope. Within days—within hours in some cases—they suddenly flashed, then faded. The rest of the time they stayed quiescent at their lowest ranges. One needed great good luck or a series of photographic snapshots to glimpse their brief brightening. Just such a series of pictures taken within two to three days of each other in 1905 had alerted Miss Leavitt to the untold number of quick-change stars in the Magellanic Clouds.
• • •
MISS LEAVITT HAD BEEN CALLED HOME AGAIN to Wisconsin after her father died on March 4, 1911, and she spent the spring and summer helping her mother settle Reverend Leavitt’s small estate. Back in Cambridge in the fall, she found the observatory family still adjusting to the loss of Mrs. Fleming. Miss Cannon was supervising the computers. Mabel Gill, a staff member since 1892, had taken over the preparation of several Annals volumes for the printer, and, together with another experienced coworker, Sarah Breslin, was wrapping up Mrs. Fleming’s longstanding effort to measure the variables she had discovered against the 222 star sequences she tailor-made for that purpose. Miss Maury had fled once more to the old Draper homestead at Hastings-on-Hudson.
Resuming the hunt for new variable stars on the Harvard sky maps, Miss Leavitt continued to ponder the thousands she had encountered in the Magellanic Clouds.
The prevalence of variables in those two southern star Clouds beggared all comparison. Miss Leavitt had tallied more than nine hundred in the Small Cloud and eight hundred in the Large, without even venturing into the Clouds’ centers, where stars clumped inseparably together.
“If the stars were equally dense over the whole sky,” Solon Bailey guessed, “their number would exceed ten billions, and the sky would be so luminous that there would be no real night.” Bailey had scanned the southern sky at Arequipa, from the decks of ships in the south latitudes, and on the Great Karroo of South Africa. In the perfect darkness of those remote sites, he had seen the star-spangled Milky Way spill across the night’s horizons. His telescopes had collapsed the vast distances to immerse him in that river of stars. Miss Leavitt, denied such intimacy with the heavens, could only imagine herself standing agape in the Andes, under the southern meanders of the Milky Way, watching the Magellanic Clouds trail after the star stream like a pair of lost sheep.
Bailey believed the two Clouds to be unique structures, separate from the Milky Way. If so, if they in fact existed outside the bounds of the galaxy, then each Cloud constituted its own so-called island universe. Possibly the numerous other splotchy white nebulous objects scattered through space were also separate star systems, independent of the Milky Way.
Bailey’s two- and four-hour exposures of the Magellanic Clouds, taken with the Bruce telescope, had revealed crowds of stars as faint as seventeenth magnitude. Miss Leavitt picked her way among them in her initial study by repeating the Baileys’ globular-cluster strategy: She ruled a reticule of one-centimeter squares on a glass plate, rendering it a transparent sheet of graph paper. Then, superposing the reticule on images of the Clouds, she cordoned off small groups of stars and charted them through an eyepiece fitted with micrometrical crosshairs.
Immune to all distractions, she differentiated the individual members, numbered them, recorded their relative positions, and tracked the variables’ brightness changes through time. The proximity of variables to one another complicated her task, as did their distances from suitable comparison stars. The pattern of the variables’ alteration also challenged her, since most of them remained at their dimmest most of the time, brightening suddenly in short bursts. In her 1908 publication, “1777 Variables in the Magellanic Clouds,” she gauged all the ranges of magnitude and gave the maximum and minimum value for every star, as best she could. She traced the complete light curves of only sixteen stars, yet this small, select sample (one one-hundredth of the whole) had shown an intriguing trend: the brighter variables had the longer periods, as though the one thing depended on the other.
Since the sixteen variables all belonged to the compact Small Cloud, Miss Leavitt reasoned they all lay roughly the same distance from Earth, just as all her relatives in Beloit lived about equally far from Cambridge. Therefore the ones that looked brighter must actually be brighter.
The unexpected correlation of brightness with period could be mere coincidence, as Miss Leavitt well knew. But if the same pattern held true for a larger number of similar variables, then the correlation itself might indicate something stupendous.
In 1911 Miss Leavitt tracked the step-by-step variation in another nine stars on the glass plates. As before, the brighter variables took the longest times to cycle through their variations. She plotted the numbers on a graph, with period lengths along the x axis, maximum and minimum magnitudes on the y. Connecting the dots, she got two smooth curves, and when she reduced the periods to a logarithmic scale, her curves snapped into straight lines. The trend among Miss Leavitt’s stars was real. Pickering called it “remarkable” when he announced her results in a Harvard College Observatory Circular on March 3, 1912. He used the word “law” to describe the finding she had demonstrated for twenty-five stars in the Small Magellanic Cloud: The brighter the magnitude, the longer the period. It meant that certain types of variables telegraphed their true magnitudes in the duration of their light cycles. Those stars heralded the coming of distance markers in the farther reaches of space. As soon as astronomers learned the key to the stellar code—the degree of brightness linked to e
ach period—they could determine the stellar magnitudes by watching a clock, then leap the interstellar distances on the wings of Isaac Newton’s inverse square law: a variable only one-quarter as bright as another of the same period must lie twice as far away.
Ejnar Hertzsprung in Denmark seized on Miss Leavitt’s period-luminosity relation. He, too, had been drawing graphs, plotting one stellar characteristic against another to test their interdependence. Like many but not all of his contemporaries, Hertzsprung saw the Draper spectral classification as a temperature gradient: The blue-white O stars were the hottest, the red M the coolest. Two red stars of nearly identical spectra therefore shared the same temperature; if one of them looked brighter than the other, then it must be either closer or larger. Hertzsprung could often judge relative distances between two such stars by their proper motion. If the farther one—the star that moved less—was the brighter of the two, then it necessarily had more surface area from which to radiate its light. This reasoning had opened Hertzsprung’s mind to the possibility of exceptionally large stars, or giants. In the past he had lauded Miss Maury for noticing the spectral nuance that could separate giants from dwarfs. Now he thanked Miss Leavitt for the means to measure distances previously beyond reach.
Hertzsprung identified about a dozen examples of stars like Miss Leavitt’s within the Milky Way. They followed the same kind of light curve, with a steep rise to peak brightness giving way to gradual decline. These stars shone many orders of magnitude brighter than their peers of the same period among Miss Leavitt’s stars. The differences put the Small Magellanic Cloud, by Hertzsprung’s reckoning, at a distance of thirty thousand light-years—a chasm so great as to strain credulity.
