When Computers Were Human

by David Alan Grier

  When the American Nautical Almanac began operations in July 1849, the navy appointed Charles Henry Davis as the first superintendent. In turn, Davis selected Peirce as the chief mathematician and established the almanac office in Cambridge, Massachusetts.19 When called to explain why he did not place the office in Washington, Davis wrote that Cambridge had “the best scientific libraries of the country—an indispensable aid in laying the permanent foundation of a work of this magnitude and importance.”20 Davis may have been rationalizing a decision that was convenient for himself and for Peirce, but he clearly wanted to create a superior publication with a staff of “first class computers.” There would be no former hairdressers on his staff and no boy computers like those found at Airy’s observatory. He stated that his computers “must be gentlemen of liberal education and of special attainments in the science of astronomy.”21

  At the start, Davis tried to recruit mathematicians for the almanac staff. He wrote to professors at the University of Virginia, Princeton, Rutgers, Columbia, and the Military Academy at West Point, asking each to serve on the almanac staff.22 To this list he added the name of one foreigner, the French astronomer Urban Jean Joseph Le Verrier (1811–1877). In 1849, Le Verrier was basking in the fame of having discovered the planet Neptune, an event that dominated the public imagination of the 1840s. He was a mathematical astronomer, not an observer, and his discovery was an event of “pure calculation,” “the grand triumph of celestial mechanics as founded on Newton’s law of gravitation.”23 Le Verrier had estimated the location of the planet by analyzing variations in the motion of Jupiter, Saturn, and Uranus and by hypothesizing the existence of an undetected body in orbit around the Sun. He had projected the possible location of this planet and sent his estimate to the observatory at Berlin. The German observers had found the planet with a single night of searching. They waited a second night in order to confirm their observation and then wrote to Le Verrier, “Monsieur, the planet, of which you indicated the position, really exists.”24

  The discovery immediately pushed Le Verrier into public view. Many urged that the planet should be named Le Verrier, just as the comet of 1682 had been named for Edmund Halley. Observers spent their nights studying the barely perceptible motion of the planet and looking for evidence of moons. The news penetrated so deeply into public consciousness that it even reached the backwoods of Massachusetts, where it received a disparaging response from Henry David Thoreau (1817–1862) at Walden Pond. In the manuscript of Walden, Thoreau dismissed the mathematics that might allow an astronomer “to discover new satellites to Neptune” but “not detect the motes in his eyes, or to what vagabond he is a satellite himself.”25 Such words failed to touch Le Verrier’s reputation. Had he been willing to join the American almanac staff, Le Verrier would have brought substantial fame and prestige to the periodical, but he declined Davis’s request. Davis was ultimately rejected by all of the mathematics professors on his list, save for Benjamin Peirce.26

  To build his computing staff, Davis turned to students, independent astronomers, and skilled amateurs. These individuals were not true members of the country’s small scientific community, but each of them had the promise of a successful future. Most of them were either friends or former students of Benjamin Peirce. The list included a young mathematical prodigy from Vermont, Henry Safford (1836–1901); a future president of the Massachusetts Institute of Technology, John Runkle (1822–1902); and a professor, Joseph Winlock (1826–1875) of Shelby College in Kentucky. Though Davis had not been able to convince the discoverer of Neptune to join the almanac staff, he was at least able to recruit an astronomer who had achieved some fame by computing Neptune’s orbit, Sears Cook Walker (1805–1853).27

  In the fall of 1846, the fall when Le Verrier announced the existence of Neptune, Walker was an assistant astronomer at the newly opened Naval Observatory in Washington. This was not a job that he had intended to take, but he had little choice in the matter. For nearly twenty years, he had worked for the Pennsylvania Company for Insurances on Lives and Granting Annuities. By day, he was an actuary, a mathematician responsible for estimating the risk and profit of insurance policies. By night, he was an amateur astronomer, using the telescopes of the Philadelphia High School to study the moon as it passed in front of stars.28 At the time, the high school observatory was the best equipped in the nation. Most of the instruments had been donated by Walker, who was making a substantial fortune in his actuarial practice. This life had come to an end in 1845, when a “series of unfortunate investments and commercial operations led to most disastrous results.” According to a friend, this disaster left Walker “at the age of forty years utterly without means.”29

  Under the best of circumstances, Sears Cook Walker could be difficult. “I have had some differences with him,” wrote Benjamin Peirce, “but they have not blinded me to his great merits.”30 At the Naval Observatory, Walker showed his great skill and his wayward nature. In the fall of 1846, he turned from his observatory duties to consider the planet Neptune. The observatory had been the recipient of small packets of data from the European astronomers, sealed with wax and addressed to the new “National Observatory.” By late fall, he had a substantial collection of data that showed the planet’s slow march across the celestial sphere. Walker decided that he might get a better calculation of the orbit if he could find an earlier observation of the planet, an observation that had been falsely recorded as a star. Using the data he had collected, he spent about three months computing the motion of the planet backwards from its location in 1846. He finally found his prediscovery observation in a 1795 star catalog that had been compiled by Joseph Lalande. The catalog showed a star where Neptune should have been. Using the navy’s telescopes, Walker sought in vain for the star and concluded that Lalande had seen the moving planet instead of a fixed star. With Lalande’s observation, Walker was able to make a refined calculation and show that the planet moved in an orbit that was nearly circular.31 The navy treated Walker’s work as a major triumph, the first important accomplishment of its observatory. “The theory of Neptune belongs, by the right of precedent, to American science,” bragged the observatory director.32

  Ironically, the most well-known of the first computers for the American Nautical Almanac was the sole woman, Maria Mitchell (1818–1889). When Charles Henry Davis appointed her to his staff, he presented her credentials to the secretary of the navy as “the lady who lately received from the King of Denmark a medal for the discovery of a new comet.”33 As an astronomical discovery, the comet was small and unimpressive. It was not even a periodic comet like Halley’s. Once it vanished from the night sky, it was gone forever. Still, the discovery caught the public’s attention as “one of the first additions to science” made in the United States.34 Mitchell was an extraordinary scientific talent, and she lived in one of the few communities that acknowledged a public role for women in nineteenth-century America. There were no other female scientists in the United States and few women in any other field of endeavor. She had discovered her comet in 1847, the year before the 1848 Seneca Falls Conference on the Rights of Women, the event that has often been identified as the start of the American feminist movement. Behind the Seneca Falls Conference lay the Quaker Church, the only religious denomination of the time that allowed women to preach to its congregations. Four of the five conference organizers were Quakers, as were Maria Mitchell and her father, William Mitchell (1791–1869). William Mitchell supported and pushed his daughter in her career, demonstrating what the early feminist Margaret Fuller (1810–1850) called the “chance of liberality” that a father might show to his daughter but a “man of the world” might never show toward his wife.35

  William Mitchell was a banker on Nantucket Island and an amateur astronomer. Maria Mitchell recalled that her interest in astronomy was “seconded by my sympathy with my father’s love for astronomical observation.”36 He taught her how to use a telescope, how to reduce data, and how to compute time from the position of the m
oon. He was a friend of the scientists at Cambridge, including Benjamin Peirce, Charles Henry Davis, and the director of the Harvard Observatory. When his daughter told him that she had identified a comet, he encouraged her to notify the observatory. When she resisted the idea, William Mitchell communicated the discovery himself. “Maria discovered a telescopic comet at half past ten on the evening of the first instant,” he wrote. “Pray tell me whether it is one of George’s [Airy of the Greenwich Observatory]. … Maria’s supposed it may be an old story.” The observatory confirmed that the comet was new and initiated an energetic effort to ensure that Maria Mitchell was given credit for the discovery. The object of their effort was King Christian VIII of Denmark. The king awarded a medal to the discoverer of any comet. In the early winter of 1848, he was preparing to give his medal for Mitchell’s comet to an Italian astronomer. The director of the Harvard Observatory did everything he could to convince the king to change his mind. He enlisted the aid of the Harvard president and the American consul in Copenhagen. Eventually, the king relented and recognized Mitchell as the first to see the comet.37

  Charles Henry Davis recruited computers through the end of 1849 and began operations that winter. He organized the staff after the pattern that Nevil Maskelyne had established some eighty years before. Each computer took responsibility for one or two tables. Sears Cook Walker took the computations for Neptune. Benjamin Peirce handled the ephemeris of Mars and the apparent movement of the Sun. Without a trace of irony, Davis asked Mitchell to handle the computations for the planet Venus. “As it is ‘Venus who brings everything that’s fair,’” he wrote, “I therefore assign you the ephemeris of Venus, you being my only fair assistant.”38 John Runkle, the future president of the Massachusetts Institute of Technology, prepared part of the calculations of lunar motion, splitting the task with another computer, just as Maskelyne’s computers had done.

  10. Maria Mitchell

  For about half of the computers, including Maria Mitchell and Sears Cook Walker, Charles Henry Davis distributed detailed computing plans through the mail. Unlike Maskelyne, Davis did not prepare hand-drawn computing forms. Though he gave the computers a rough idea of how the sheet should look, he let them organize the computations as they desired. “You may fill up the sheets as much as possible, consistently with the clearness,” he told Maria Mitchell, and in so doing “can thus economize on paper.”39 He provided paper and reference books to all of the computers, sending the supplies with a private forwarding company called Adams Express. Mitchell, who had a relatively meager library, often requested books for her work. In December 1849, she asked for a copy of Theoria Motus Corporum Coelestium in Sectionibus Conicis Solem Ambientium (The Theory of the Motion of the Heavenly Bodies Moving about the Sun in Conic Sections) by the German astronomer Carl Friedrich Gauss (1777–1855). This book, one of the last major astronomical texts written in Latin, had a fairly complete discussion of the techniques of astronomical calculations. “I have directed my bookseller to endeavor to get two copies [for the almanac office],” Davis responded, “and will add a third to the list for yourself if you wish it.” He concluded the letter by commenting, “I am glad you read Latin.”40

  The remaining computers, those who lived near Cambridge, worked at the almanac office. As far as we can determine, this office resembled the rooms of the British Nautical Almanac in London. There were worktables for the computers and a private area for the superintendent. One young computer recalled coming into the Cambridge office on a frosty January morning, taking a “seat between two well-known mathematicians, before a blazing fire.” It was an informal place, where new ideas of mathematics were freely discussed between calculations, as the “discipline of the public service was less rigid in the office at that time than at any government institution I ever heard of.” Each computer was expected to spend five hours a day in the office. “The hours might be selected by himself, and they generally extended from nine until two, the latter being at that time the college and family dinner hour.” All that Davis required was that the work was done on time.41

  The disjointed operation of the almanac, with some computers in Cambridge and some working from home, fit easily into the organizational structure of the navy. Davis kept track of each computer and the progress of his or her work. Four times a year, he would send pay vouchers to the computers that could be redeemed at any naval facility. The letters that accompanied these payments contain gossip, discussions of mathematics, news about astronomy, and even a few quotes from Shakespeare, who was clearly his favorite author. “Enclosed are your vouchers signed by myself,” he wrote to a new computer; “you had better negotiate them with a friendly broker, rather than with one where you hold a relation of Antonio to Shylock.”42 Almanac computers earned between five hundred and eight hundred dollars per year for their work. Two of them doubled their income by checking and correcting the work of others in addition to doing their own computations.43

  After a year of calculations, Davis could report that work for the first issue, which would cover the year 1855, was progressing nicely. He told the secretary of the navy that his “small corps of computers” was employing the theories of the “illustrious Leverrier,” using the corrections of Airy, and reducing data from Maskelyne. He reminded the secretary that he had already made a special report of “a variation in the proper motion of one of the fundamental stars,” which he claimed “has led to a discovery of particular interest in stellar astronomy.” He summarized his progress by stating, “I have frequently expressed my wish that the Nautical Almanac should in every respect conform to the most advanced state of modern science and be honorable to the country and it is my determination to spare no effort by which this high object can be attained.”44

  Though Davis was confident that his computing staff would produce an American almanac that rivaled the publications of Europe, others were not so certain. The potential users of the almanac, notably the merchant navigators and the surveyors, were willing to withhold their judgment until they saw the final product. These two groups were concerned about a basic element of the almanac, the location of the prime meridian. All almanacs had to establish a meridian, a line running through the north and south poles that would serve as a reference for the positions of stars. The Connaissance des Temps used the Paris meridian, which passed through the Cathedral of Notre Dame. The British Nautical Almanac was prepared for the Greenwich meridian. Originally, this line had been drawn through the Octagon Room of the Royal Observatory, but it had since moved to the site of a new telescope mount a few yards to the west. The original proposal for the American Ephemeris and Nautical Almanac instructed the navy to publish an “American Nautical Almanac, to be calculated for the meridian of Washington city.”45

  Davis did not like the idea of a Washington meridian. He preferred to use the Greenwich line, for he felt that it would produce the most accurate longitude calculations. “Our own vessels are constantly meeting those of Great Britain on the great highway of nations,” he noted, “and are in the habit of comparing with them their longitudes.”46 By adopting the Greenwich meridian, the computers of the American almanac could use without additional calculations the vast catalogs of star data that George Airy was compiling at the Royal Observatory. However, the Greenwich meridian did not satisfy the computers of the Coast Survey or the surveys of the various American states. They wanted a meridian safely on the North American continent. Their calculations would be most accurate if they could physically measure parts of the meridian and if they could work without an ocean intervening between themselves and their baseline.

  From one point of view, the argument between those favoring the Greenwich meridian and those who wanted a Washington meridian was an honest scientific disagreement. Each of the two groups advocated a procedure that was best for its own needs. Each of the procedures would provide an acceptable, though not perfect, solution for the other group. Yet the argument had symbolic and economic aspects beyond the scientific construction of astron
omical tables. The symbolic problem was an offshoot of Manifest Destiny and the Monroe Doctrine. The American citizenry saw themselves settling the breadth of the North American continent, from the Atlantic to the Pacific. As settlers moved west, they would use the American Nautical Almanac to survey the territories and set the borders of new states. Congress would insist that those borders be specified from Washington, rather than from a meridian that passed through the suburbs of a foreign capital.

  Davis believed that he had a clever compromise to the meridian problem, one that would satisfy both navigators and surveyors. He proposed “to establish an arbitrary meridian at the city of New Orleans, which will be exactly six hours in time, or ninety degrees in space, from the meridian of Greenwich.”47 The idea was an elegant technical solution, as it gave the United States its own meridian and allowed navigators to compute their position relative to Greenwich with little extra effort. Davis must have believed that his idea would be accepted without complaint, but just as he was starting the computations, he reported to the navy “that a ‘remonstrance’ against my paper on the American Prime Meridian and against any change from the Meridian of Greenwich is circulated among the merchants and insurers of the cities of Boston and New York for signatures.”48