The Glass Universe Read online

  As arranged, she reached Summerhouse Hill above Harvard Yard on Friday morning, February 9, accompanied by her husband’s close friend and colleague George F. Barker of the University of Pennsylvania. Barker, who was preparing a biographical memoir of Henry, had been the Drapers’ houseguest at the time of the Academy dinner. Late that night, when Henry was seized with a violent chill while bathing, it was Barker who helped lift him from the tub and carry him to the bedroom. Then he bid the Drapers’ neighbor and physician Dr. Metcalfe, another dinner guest, to return to the house immediately. Dr. Metcalfe diagnosed double pleurisy. Although Henry of course received the most tender nursing—and showed some brief promise of improvement—the infection spread to his heart. On Sunday the doctor noted the signs of pericarditis, which precipitated Henry’s death at about four o’clock Monday morning, the twentieth of November.

  • • •

  MRS. DRAPER HAD VISITED OBSERVATORIES with her husband in Europe and the States, but she had not set foot inside one in months. At Harvard, the large domed building that housed the several telescopes doubled as the director’s residence. Both Professor and Mrs. Pickering ushered her into the pleasant rooms and made her feel welcome.

  Mrs. Pickering, née Lizzie Wadsworth Sparks, daughter of former Harvard president Jared Sparks, did not aid her husband in his observations, as Mrs. Draper had done, but acted as the institution’s vivacious and charming hostess.

  An exaggerated though genuine politeness characterized the directorial style of Edward Charles Pickering. If the observatory’s financial straits constrained him to pay his eager young assistants meager wages, nothing prevented his addressing them respectfully as Mr. Wendell or Mr. Cutler. He called the senior astronomers Professor Rogers and Professor Searle, and all but doffed his hat and bowed to the ladies—Miss Saunders, Mrs. Fleming, Miss Farrar, and the rest—who arrived each morning to perform the necessary calculations upon the nighttime observations.

  Was it usual, Mrs. Draper wondered, to employ women as computers? No, Pickering told her, as far as he knew the practice was unique to Harvard, which currently retained six female computers. While it would be unseemly, Pickering conceded, to subject a lady to the fatigue, not to mention the cold in winter, of telescope observing, women with a knack for figures could be accommodated in the computing room, where they did credit to the profession. Selina Bond, for example, was the daughter of the observatory’s revered first director, William Cranch Bond, and also the sister of his equally revered successor, George Phillips Bond. She was currently assisting Professor William Rogers in fixing the exact positions (in the celestial equivalents of latitude and longitude) for the several thousand stars in Harvard’s zone of the heavens, as part of a worldwide stellar mapping project administered by the Astronomische Gesellschaft in Germany. Professor Rogers spent every clear night at the large transit instrument, noting the times individual stars crossed the spider threads in the eyepiece. Since air—even clear air—bent the paths of light waves, shifting the stars’ apparent positions, Miss Bond applied the mathematical formula that corrected Professor Rogers’s notations for atmospheric effects. She used additional formulas and tables to account for other influential factors, such as Earth’s progress in its annual orbit, the direction of its travel, and the wobble of its axis.

  Anna Winlock, like Miss Bond, had grown up at the observatory. She was the eldest child of its inventive third director, Joseph Winlock, Pickering’s immediate predecessor. Winlock had died of a sudden illness in June 1875, the week of Anna’s graduation from Cambridge High School. She went to work soon afterward as a computer to help support her mother and younger siblings.

  Williamina Fleming, in contrast, could claim no familial or collegiate connection to the observatory. She had been hired in 1879, on the residence side, as a second maid. Although she had taught school in her native Scotland, certain circumstances—her marriage to James Orr Fleming, her immigration to America, her husband’s abrupt disappearance from her life—forced her to seek employment in a “delicate condition.” When Mrs. Pickering recognized the new servant’s abilities, Mr. Pickering reassigned her as a part-time copyist and computer in the other wing of the building. No sooner had Mrs. Fleming mastered her tasks in the observatory than the impending birth of her baby sent her home to Dundee. She stayed there more than a year after her confinement, then returned to Harvard in 1881, having left her son, Edward Charles Pickering Fleming, in the care of her mother and grandmother.

  • • •

  NONE OF THE PROJECTS UNDER WAY at the observatory looked familiar to Mrs. Draper. Henry’s amateur standing and private means had freed him to follow his own interests at the forefront of stellar photography and spectroscopy, while the professional staff here in Cambridge hewed to more traditional pursuits. They charted the heavens, monitored the orbits of planets and moons, tracked and communicated the courses of comets, and also provided time signals via telegraph to the city of Boston, six railroads, and numerous private enterprises such as the Waltham Watch Company. The work demanded both scrupulous attention to detail and a large capacity for tedium.

  When the thirty-year-old Pickering took over as director on February 1, 1877, his primary responsibility had been to raise enough money to keep the observatory solvent. It received no support from the college to pay salaries, purchase supplies, or publish the results of its labors. Aside from interest on its endowment and income from its exact-time service, the observatory depended entirely on private bequests and contributions. A decade had passed since the last solicitation for funds. Pickering soon convinced some seventy astronomy enthusiasts to pledge $50 to $200 per year for five years, and while those subscriptions trickled in, he sold the grass cuttings from the six-acre observatory grounds at a small profit. (They brought in about $30 a year, or enough to cover some 120 hours’ worth of computing time.)

  Born and bred on Beacon Hill, Pickering navigated easily between the moneyed Boston aristocracy and the academic halls of Harvard University. In his ten years spent teaching physics at the fledgling Massachusetts Institute of Technology, he had revolutionized instruction by setting up a laboratory where students learned to think for themselves while solving problems through experiments that he designed. Pursuing his own research at the same time, he explored the nature of light. He also built and demonstrated, in 1870, a device that transmitted sound by electricity—a device identical in principle to the one perfected and patented six years later by Alexander Graham Bell. Pickering, however, never thought to patent any of his inventions because he believed scientists should share ideas freely.

  At Harvard, Pickering chose a research focus of fundamental importance that had been neglected at most other observatories: photometry, or the measurement of the brightness of individual stars.

  Obvious contrasts in brightness challenged astronomers to explain why some stars outshone others. Just as they ranged in color, stars apparently came in a range of sizes, and existed at different distances from Earth. Ancient astronomers had sorted them along a continuum, from the brightest of “first magnitude” down to “sixth magnitude” at the limit of naked-eye perception. In 1610 Galileo’s telescope revealed a host of stars never seen before, pushing the lower limit of the brightness scale to tenth magnitude. By the 1880s, large telescopes the likes of Harvard’s Great Refractor could detect stars as faint as fourteenth magnitude. In the absence of uniform scales or standards, however, all estimations of magnitude remained the judgment calls of individual astronomers. Brightness, like beauty, was defined in the eye of the beholder.

  Pickering sought to set photometry on a sound new basis of precision that could be adopted by anyone. He began by choosing one brightness scale among the several currently in use—that of English astronomer Norman Pogson, who calibrated the ancients’ star grades by presuming first-magnitude stars to be precisely a hundredfold brighter than those of sixth. That way, each step in magnitude differed from the next by a factor of 2.
512.

  Pickering then chose a lone star—Polaris, the so-called polestar or North Star—as the basis for all comparisons. Some of his predecessors in the 1860s had gauged starlight in relation to the flame of a kerosene lamp viewed through a pinhole, which struck Pickering as tantamount to comparing apples with oranges. Polaris, though not the sky’s brightest star, was thought to give an unwavering light. It also remained fixed in space above Earth’s north pole, at the hub of heavenly rotation, where its appearance was least susceptible to distortion by currents of intervening air.

  With Pogson’s scale and Polaris as his guides, Pickering devised a series of experimental instruments, or photometers, for measuring brightness. The firm of Alvan Clark & Sons built some dozen of Pickering’s designs. The early ones attached to the Great Refractor—the observatory’s premier telescope, a gift from the local citizenry in 1847. Ultimately Pickering and the Clarks constructed a superior freestanding model they called the meridian photometer. A dual telescope, it combined two objective lenses mounted side by side in the same long tube. The tube remained stationary, so that no time was lost in repointing it during an observing session. A pair of rotating reflective prisms brought Polaris into view through one lens and a target star through the other. The observer at the eyepiece, usually Pickering, turned a numbered dial controlling other prisms inside the instrument, and thus adjusted the two lights until Polaris and the target appeared equally bright. A second observer, most often Arthur Searle or Oliver Wendell, read the dial setting and recorded it in a notebook. The pair repeated the procedure four times per star, for several hundred stars per night, exchanging places every hour to avoid making errors due to eye fatigue. In the morning they turned over the notebook to Miss Nettie Farrar, one of the computers, for tabulation. Taking Polaris’s arbitrarily assigned magnitude of 2.1 as her base, Miss Farrar arrived at relative values for the other stars, averaged and corrected to two decimal places. By these means, it took three years for Pickering and his crew to pin a magnitude on every star visible from the latitude of Cambridge.

  The objects of Pickering’s photometry studies included some two hundred stars known to vary their light output over time. These variable stars, or “variables,” required the closest surveillance. In his 1882 report to Harvard president Charles Eliot, Pickering noted that thousands of observations were needed to establish the light cycle of any given variable. In one instance, “900 measures were made in a single night, extending without intermission from 7 o’clock in the evening until the variable had attained its full brightness, at half past 2 in the morning.”

  Pickering needed reinforcements to keep watch over the variables. Alas, in 1882, he could not afford to hire even one additional staff member. Rather than dun the observatory’s loyal subscribers for more money, he issued a plea for volunteers from the ranks of amateur observers. He believed women could conduct the work as well as men: “Many ladies are interested in astronomy and own telescopes, but with two or three noteworthy exceptions their contributions to the science have been almost nothing. Many of them have the time and inclination for such work, and especially among the graduates of women’s colleges are many who have had abundant training to make excellent observers. As the work may be done at home, even from an open window, provided the room has the temperature of the outer air, there seems to be no reason why they should not thus make an advantageous use of their skill.”

  Pickering felt, furthermore, that participating in astronomical research would improve women’s social standing and justify the current proliferation of women’s colleges: “The criticism is often made by the opponents of the higher education of women that, while they are capable of following others as far as men can, they originate almost nothing, so that human knowledge is not advanced by their work. This reproach would be well answered could we point to a long series of such observations as are detailed below, made by women observers.”

  Pickering printed and distributed hundreds of copies of this open invitation, and also convinced the editors of several newspapers to publish it. Two early responses arrived in December 1882 from Eliza Crane and Mary Stockwell at Vassar College in Poughkeepsie, New York, followed by another from Sarah Wentworth of Danvers, Massachusetts. Pickering began assigning particular variables to individuals for observation. Although his volunteers lacked any equipment as sophisticated as the meridian photometer, they could compare their variables with other nearby stars, and estimate the brightness changes over time. “If any of the stars become too faint,” he advised them by letter, “please send word, so that observations may be attempted here” with the large telescope.

  Some women wrote to request formal instruction in practical or theoretical astronomy, but the observatory provided no such courses, nor could it admit curious spectators, male or female, at night. During the day, the director would be only too pleased to show visitors around the building.

  Pickering’s daytime duties as director required him to correspond regularly with other astronomers, purchase books and journals for the observatory’s library, attend scientific meetings, edit and publish the Annals of the Astronomical Observatory of Harvard College, oversee finances, answer inquiries by mail from the general public, host visiting dignitaries, and order supplies large and small, from telescope parts to furnace coal, stationery, pens, ledgers, even “water closet paper.” Every bit of observatory business demanded his personal attention or at the very least his signature. Only when a blanket of clouds hid the stars could he find a night’s sleep.

  • • •

  MRS. DRAPER’S GLASS PLATES demanded examination by daylight. Although Pickering had heard much about these images, and even discussed them with the doctor the night of the Academy dinner in November, he had not seen them till now. He was accustomed to looking at spectra—the separated rays of starlight—through the telescope, using attachments called spectroscopes that former director Joseph Winlock had purchased in the 1860s, when spectroscopy came into vogue. The live view through the spectroscope turned a star into a pale strip of colored light ranging from reddish at one end through orange, yellow, green, and blue to violet at the other. The spectroscope also made visible many black vertical lines interspersed at intervals along the colored strip. Astronomers believed that the breadth, intensity, and spacing of these spectral lines encoded vital information. Though the code remained unbroken, a few investigators had proposed schemes to classify the stars by type, according to the similarities in their spectral line patterns.

  On the Draper plates, each spectrum looked like a gray smudge barely half an inch long, yet some contained as many as twenty-five lines. As Pickering viewed them under a microscope, their detail stupefied him. What skill their capture demonstrated, and what luck! He knew of only one other person in the world—Professor William Huggins of England—who had ever succeeded in capturing a stellar spectrum on a photographic plate. Huggins was also the only man of Pickering’s acquaintance, aside from Dr. Draper, to have discovered an able astronomical assistant in his own wife, Margaret Lindsay Huggins.

  Mrs. Draper agreed to leave her plates in Pickering’s care for a complete analysis, and returned to New York. She promised Mrs. Pickering, who was considered one of Cambridge’s most accomplished gardeners, to visit again in spring or summer, in the hope of seeing the observatory grounds in full bloom.

  Pickering measured each spectrum with a screw-thread micrometer. By February 18, 1883, he could report to Mrs. Draper that he was finding “much more in the photographs than appears at first sight.” The computers had plenty to do in graphing the readings from his every half-turn of the screw, then applying a formula and computations to translate them into wavelengths. It became clear that Dr. Draper had demonstrated the feasibility of studying the stellar spectra by means of photography, instead of by peering through instruments and drawing a record of what the eye saw.

  Pickering again pressed Mrs. Draper to publish an illustrated account, not merely to
establish priority for her husband, but, more important, to show other astronomers the great promise of his technique.

  For help with the preparation of the paper, Mrs. Draper asked a noted authority on the solar spectrum, Charles A. Young of Princeton, to contribute an introduction outlining Henry’s methods. Meanwhile she catalogued all seventy-eight plates in the spectra series, relying on Henry’s notebooks to specify the date and time of each photograph taken, the star name, the length of every exposure, the telescope used, and the width of the spectroscope slit, plus incidental remarks about observing conditions, such as “There was blue fog in the sky” or “The night was so windy that the dome was blown around.”

  Pickering summarized the twenty-one plates he had scrutinized in ten tables with explanations. He reported the distances between spectral lines, stating the methodology and mathematical formulas employed to translate line positions into wavelengths of light. He also commented on the similar work being done by William Huggins in London, and ventured to categorize some of Draper’s spectra by Huggins’s criteria. When he sent his draft to Mrs. Draper for approval, she balked at the mention of Huggins.

  “Dr. Draper did not agree with Dr. Huggins,” she wrote Pickering on April 3, 1883, concerning two of the stars in the series. Their nearly identical spectra both showed wide bands, which had made Huggins classify the two stars as a single type, but the Draper photographs revealed that one of these stars also had many fine lines between the bands, which set it apart from the other. “In view of this I should not like to accept Mr. Huggins’ classification as the standard when Dr. Draper did not agree with it.” Although Pickering had seen the abundance of fine lines she described, he found them too delicate for satisfactory measurement.