Finding Longitude
Page 9
CHAPTER 4
MAKING LONGITUDE WORK
We have employed proper persons to compute a Nautical Almanac and Astronomical Ephemeris ... which will greatly contribute to make the Lunar Tables constructed by the late Professor Mayer ... more generally useful ...
Foreword to The Nautical Almanac and Astronomical Ephemeris, for the Year 17671
The 1765 Longitude Act had two main functions. First, it provided a statutory process through which to reward key individuals, such as John Harrison, who had developed successful ways to measure longitude. Second, it provided a way of investing in and supporting, for the foreseeable future, the two methods of finding longitude that had proved to be the most successful: lunar-distance and timekeeper. This strategy was not a way of hedging bets, or of supporting one method until the other was available for wider use, but was based on the view that they were complementary. Astronomical methods, though tricky to use, were the only means of finding longitude if it were lost and of checking that a timekeeper was working sufficiently well. They were also considered the more accurate methods, under ideal observing conditions and with many repetitions.
The new Act also recognized something of which the Commissioners had become ever more aware: the terms of the 1714 Act were insufficient for selecting a method that could simply be rolled out across the Navy. The Commissioners were conscious that they were dispensing large amounts of public money and were at risk of having little to show for it. The need for further work and investment had to be explained carefully to Members of Parliament, who might baulk at continued spending and fail to grasp why a successful trial did not mean the problem was solved.
For the lunar-distance method, successful implementation meant continually publishing predictive astronomical tables to ease the burden of calculation for the user. The 1765 Act therefore stated that to make Mayer’s ‘Lunar Tables more generally useful’, the Commissioners should
cause such Nautical Almanacks, or other useful Tables, to be constructed, and to print, publish, and vend ... any Nautical Almanack or Almanacks, or other useful Table or Tables, which they ... shall, from time to time, judge necessary and useful, in order to facilitate the Method of discovering the Longitude at Sea ...2
* * *
The bare construction of one single machine can never be said to discover the longitude.
* * *
From an anonymous letter to The Gentleman’s Magazine (1765)3
Fig. 1 – The Royal Observatory from the south-east, unknown artist, c.1770
{National Maritime Museum, Greenwich, London}
Nevil Maskelyne, now Astronomer Royal, was in a position to focus work at the Royal Observatory on the production of observations that would serve this purpose. On top of this, he directed the considerable calculating work required to turn raw observational data into usable tables, published as the Nautical Almanac. This was, as suggested in the British Mariner’s Guide, to be modelled on the tables for 1761 that had been computed by the French astronomer Nicolas Louis de Lacaille for the Connaissance des Temps.
The Almanac also supported the use of timekeepers at sea. Anticipating a future in which they were more plentiful, it provided information on how to take and process observations to establish local time, compare this to a timekeeper and check its going (that is, its speed and regularity), at sea or on land. However, there was much more to be done to make this method widely available. One watch – one very expensive, complex and slowly constructed watch – was of little use: the Commissioners had to know that Harrison’s work could be replicated. Harrison was, after all, now in his seventies and it was a very real possibility that his secrets might die with him.
Fig. 2 – The title page and one of the tables from The Nautical Almanac and Astronomical Ephemeris for 1767
{National Maritime Museum, Greenwich, London}
By this time, reflecting their more regular transaction of business, the Commissioners had become widely known as the Board of Longitude. Their collective decision, at a meeting on 9 February 1765, was that, although the watch had passed the trial, they could not be confident that it had solved the problem. Harrison would only be paid ‘upon his producing his Timekeeper to certain persons to be named by this Board & discovering to them, upon oath, the principles & manner of making the same’.4 Harrison vehemently disagreed, believing that, since his watch had performed as specified in the 1714 Act, he should immediately receive the full reward.
Ultimately, new watches would have to be made, with cheaper and simpler constructions. Maskelyne and the Royal Observatory were central to this process. The Astronomer Royal assessed ideas, dealt with makers, encouraged the spread of horological knowledge, oversaw trials of timekeepers at the Observatory and encouraged their use at sea. He and the Board also facilitated the exchange of knowledge about longitude methods across national borders. In continental Europe, astronomers and clockmakers had been engaged in similar researches, with their own networks of patronage and financial support. And in France, in particular, effective marine timekeepers were beginning to be produced.
Maskelyne and the Nautical Almanac
Working on behalf of the Board of Longitude, Maskelyne quickly created a system for producing the Nautical Almanac and Astronomical Ephemeris (Fig. 2). Within months of his arrival at Greenwich in March 1765, he had established the layout and content of each Almanac, providing information useful to astronomy, navigation and cartography, initially for a year ahead, and later for several years in advance. Work got under way, and the data for 1767 were ready to be printed by the end of 1766. As the Astronomer Royal told his brother, then living in India,
Fig. 3 – The title page and a page of tables from the Tables Requisite to be used with the Astronomical and Nautical Almanac (London, 1766)
{National Maritime Museum, Greenwich, London}
There will be 12 pages in every month. All the lunar calculation for finding the longitude at sea by that method will be ready performed: & other useful & new tables added to facilitate the whole calculation; so that the sailers will have little more to do than to observe carefully the moon’s distance from the sun or a proper star, which are also set down in the ephemeris, in order to find their longitude.5
It was not quite as simple as this sounded but the ‘ready performed’ calculation considerably shortened the process of establishing a reference time, in this case Greenwich time. The tables would also enable the establishment of longitudes on land, by observations of Jupiter’s satellites, and facilitate calculation of latitude and local time.
Maskelyne noted also that, to complete this endeavour, ‘The board of longitude have engaged persons to compute’ this ‘very complete nautical & astronomical ephemeris’.6 These were Israel Lyons, George Witchell, William Wales and John Mapson, and they were appointed on 13 June 1765. To ensure the accuracy of the calculations, these ‘computers’, as they were known, worked independently in pairs, so that the work of each member of a pair could be checked against that of the other. Lyons and Witchell worked on the data for January to June 1767, while Wales and Mapson covered July to December. The pattern of work for the next several decades was set when Richard Dunthorne was appointed ‘comparer’ to check the accuracy of each pair’s work and correct as necessary. Each computer and comparer was paid £70 for a year’s calculations, though this was soon raised to £75, probably because the slow, painstaking nature of the work became apparent.
Dunthorne also worked with Maskelyne to produce another volume, the Tables Requisite to be used with the Astronomical and Nautical Almanac (Fig. 3), which provided additional information that did not require annual updating. It aimed to be clear and helpful, explaining how to take observations, what instruments to use and how to apply the tables. Both this and the first edition of the Almanac were available from the printer on 6 January 1767. While the Board seems to have overestimated the market – of the 1000 Almanacs and 10,000 Tables Requisite printed, 242 and 6992 respectively remained unsold by 1784 – t
his was the start of a very long series. Maskelyne oversaw and edited forty-nine issues of the Almanac and three editions of the Tables Requisite. The publication of the Almanac continues to this day.
Fig. 4 – The John Bird transit instrument and astronomical regulator made by George Graham in the Royal Observatory, Greenwich, drawing by John Charnock, about 1785 (detail)
{National Maritime Museum, Greenwich, London}
Fig. 5 – Nevil Maskelyne’s ‘observing suit’ of padded silk, probably designed to keep him warm during cold, night-time observations, about 1765
{National Maritime Museum, Greenwich, London}
Fig. 6 – Nevil Maskelyne, by Louis François Gérard van der Puyl, 1785
{The Royal Society}
Maskelyne’s dedication and management were essential to the success of the process, although he did not receive additional income for adding the role of Almanac editor to that of Astronomer Royal. Other, paid, individuals were required to sustain the project. Closest to home were the successive astronomical assistants at Greenwich. Their priority was to assist with observing the Moon, measuring its height and timing the moment that it crossed the meridian of the fixed observatory instruments, such as the John Bird transit instrument and astronomical regulator made by George Graham (Fig. 4).
The transits of the Sun and planets were also observed when possible, as were lunar and solar eclipses and other conjunctions of the Sun, Moon, planets, satellites and stars. All of these could be used to establish and refine the measurement of the difference in longitude between Greenwich and any other location at which the same observations had been made. Stellar observations were largely limited to thirty-one (later thirty-six) stars that lie near the celestial equator, the imaginary circle in the sky that lies in the same plane as the Earth’s Equator. The stars chosen were bright enough to be viewed with a telescope in daylight and were spread across the sky. These stars were used to correct the observatory clocks and became known as ‘clock stars’. An overlapping group of ten bright stars, in addition to the positions of the Sun and Moon, were observed for tabulation in the Nautical Almanac.
Observations were generally carried out by astronomer and assistant together: Maskelyne’s padded silk ‘observing suit’ shows signs of significant wear (Fig. 5). As Astronomer Royal, he worked under rules that he had helped the Royal Society establish in the year before his appointment. The rules stipulated that he should not be absent for more than a few days without permission and that he must ensure that either he or his assistant were resident at all times. It had also become a requirement for copies of the observations carried out at the Observatory to be submitted annually to its Board of Visitors. They were later printed, thus beginning the regular series of Greenwich Observations. Maskelyne’s portrait shows him with the first edition of 1776, containing ten years’ observations (Fig. 6).
It was largely Maskelyne’s assistants who processed, or reduced, the observations into data ready for publication. It was tedious and poorly paid work and seems to have led to a high turnover of staff, so Maskelyne asked that their pay be increased. The rise from £26 to £86 must have been welcome, yet, as Maskelyne himself wrote, the role still required an individual with good eyes for the telescope, good ears for the ticking of the clock, a good constitution, and the ability to work on calculations several hours a day, get up at night and ‘bear confinement’.7 The role was often seen as an opportunity for excellent observational and mathematical training rather than as a long-term position.
Of the twenty-four individuals who were assistants during Maskelyne’s forty-six year tenure, many went on to become observers elsewhere, or teachers of mathematics, astronomy and navigation. Often benefiting from Maskelyne’s patronage, they helped to spread knowledge of the new longitude methods. Several former assistants either regularly or irregularly took on computational work for the Nautical Almanac and the Board of Longitude. Such work also added to the income of others with mathematical ability, good literacy, focus and a neat hand, including teachers, astronomers, ministers, surveyors and instrument makers. Such people had useful skills but often precarious careers.
Fig. 7 – List of reference books required by the computers, compiled by Nevil Maskelyne in 1799
{Cambridge University Library}
Most of the methods they used had been outlined in the British Mariner’s Guide, although the computers and comparer undoubtedly helped refine them for production on a larger scale. Instructions were sent out by post, as were the work, queries, corrections and completed calculations. The computers were also sent a large number of astronomical and mathematical tables as reference tools (Fig. 7) that they needed to consult in order to process a set of observations. These books of tables belonged to the Board of Longitude, and Maskelyne was always careful to ensure their safe return if one of its employees ceased work.
While most of the computers were known or personally recommended to Maskelyne, they were geographically scattered. They were able to work from home and, often, in between other professional activities. Payment was organized by month of the Almanac computed, and Maskelyne became skilled at anticipating the rate of work and monetary requirements of his different computers, keeping a careful track of work as it was parcelled out (Fig. 8).
In December 1767 it was decided that the Almanac should be published three years in advance in order to provide for longer voyages. Maskelyne was authorized to recruit more computers, two of whom were Joseph Keech and Reuben Robbins. They evidently found that they had taken on more work than they anticipated, for they took the shortcut of copying each other’s calculations rather than working the same data independently. The comparer, Malachy Hitchins, proved his ability to detect errors or, in this case, suspicious regularity. Their cheating was revealed and the pair were dismissed and told to recompense Hitchins for his wasted time. Thereafter, Maskelyne made a virtue of the geographical spread of his computers, by making sure that each computer in a pair lived at a distance from each other. He obviously forgave Keech and Robbins, as they did further computing, but they never again worked on the same month’s data.
Maskelyne employed a total of thirty-five computers and comparers. By 1789, though, he was using only four – Mary Edwards, Henry Andrews, Joshua Moore and Malachy Hitchins – as the process became streamlined and the computers adept. The system and these loyal employees ensured that accuracy was largely maintained. Problems crept in after the deaths of Hitchins, who had been comparer for forty years, and Maskelyne, especially under the less strictly controlled regime of his successor, John Pond (1767–1836).
Many of the stories of individual computers make fascinating reading but Mary Edwards stands out as one of the few female contributors to this narrative. Even her name might have been lost, had she not outlived her husband. John Edwards was a Shropshire clergyman who lived near Maskelyne’s sister Margaret, Lady Clive, and probably became known to them after making his own telescope mirrors. He used his astronomical interests and mathematical skills to add significantly to his family income, receiving payment for work on six months’ worth of each Almanac from 1773 to 1784. However, it seems likely that he handed on much or most of the calculating work to his wife, for when he died the accounts moved seamlessly from ‘John Edwards’ to ‘Mary Edwards’.
Fig. 8 – List by Nevil Maskelyne of work allocated to computers and comparers in 1791–93 for the 1803 Nautical Almanac
{Cambridge University Library}
Fig. 9 – Date and signature on the upper plate of John Harrison’s H4 sea-watch
{National Maritime Museum, Greenwich, London, Ministry of Defence Art Collection}
It was unusual for women at this time to have had the mathematical knowledge and training to work as computers, but Mary may have been taught by her husband and must have had a natural aptitude. She, in turn, gained enough experience to help teach new computers, including her daughters, and after the death of Hitchins she was also able to take on work as a comparer. She worked fast and
had a very low rate of errors. Earning under her own name, she increased her workload, computing twelve months of each year’s Almanac. As half of the whole computing power required annually, this was more than any of the other computers.
After Pond took over as editor of the Nautical Almanac in 1811, Edwards suddenly found her computing work being reduced and her comparing work stopped altogether. She petitioned the Board, which acknowledged her good work but did not reinstate her to the more prestigious position of comparer. She was, however, effectively paid a premium for twelve months’ calculations while only doing eight. Edwards died in 1815 but her daughter Eliza continued as a computer until 1832, when the work of the Almanac was consolidated and ceased to be the cottage industry in which the Edwards had specialized. As in other fields in the nineteenth century, such women had their earning power removed by changes to the organization of workplaces.
The ‘Discovery of Mr Harrison’s Watch’
The Nautical Almanac was intended to help make the new longitude methods ‘practicable and useful at sea’. This phrase in the 1714 Act provided the Commissioners of Longitude with much needed wriggle-room: it was decided that only with additional information could Harrison’s watch fulfil the intent of the Act. To deal responsibly with public money, the full £20,000 reward would only be paid when the Board was satisfied ‘that his method will be of common & general Utility’.8