Finding Longitude: How ships, clocks and stars helped solve the longitude problem

This information was used to fix the current position of the ship by plotting its direction and distance travelled from one point to the next – a procedure known as ‘dead reckoning’. By applying the latest measurements to the previous day’s position, and adjusting for the effects of wind and currents, the ship’s navigator could plot the new position on the chart and note it in the written log. This was fairly straightforward over short distances. Over much longer distances, printed tables were used to convert the ship’s various diagonal courses into changes of position north–south and east–west.

Latitude could be measured directly from the maximum height of the Sun or the Pole Star above the horizon. A range of instruments for these observations had been devised over the centuries, with those in use by the late seventeenth century including the cross-staff (see Chapter 2, Fig. 14 (#litres_trial_promo)) and, particularly among English sailors, the backstaff (Fig. 9 (#litres_trial_promo)). Each measured an angle between the celestial body (usually the Sun) and the horizon, from which latitude could be derived with a few simple calculations. Until the perfection of techniques described in later chapters, however, longitude could only be derived from dead reckoning, which was indispensable on long-distance voyages. On days when the weather allowed an astronomical observation for latitude, the difference between that and the latitude calculated by dead reckoning could be used to adjust the longitude estimate, hopefully improving its accuracy.

Constant vigilance was also essential – ‘the best navigator is the best looker-out’, Samuel Pepys noted.

This included watching for additional clues to check the ship’s position, in particular when relatively close to the coast. Natural and man-made features, such as a headland, a church tower or a deliberately placed marker, were obvious signposts. As they headed ‘north up the Yorkshire coast’, for instance, Whitby sailors recalled that:

When Flamborough we pass by

Filey Brigg we mayn’t come nigh

Scarborough Castle lies out to sea,

Whitby three points northerly.

This local knowledge was also written down or published in books known as pilots or rutters (from the French routiers), which included descriptions and sketches of distinctive coastal features. When land was out of sight, birds, marine animals and plants could reveal its proximity and direction. On a voyage to Philadelphia in 1726, Benjamin Franklin was reassured that they would soon arrive, having seen

Fig. 7 – A mariner’s compass made by Jonathan Eade in London, c.1750. The compass is mounted on gimbals to keep it steady on a moving ship. North is indicated by a fleur-de-lys

{National Maritime Museum, Greenwich, London}

Fig. 8 – A page from the log of the Orford by Lieutenant Lochard, October 1707, showing the observations and results of calculations for latitude and longitude. There is also a column for general comments (detail)

{National Maritime Museum, Greenwich, London}

Fig. 9 – A backstaff, used to measure the angle between the Sun and the horizon; made of lignum vitae and boxwood by Will Garner, London, 1734

{National Maritime Museum, Greenwich, London}

Fig. 10 – A seaman with a lead and line (right), from The Great and Newly Enlarged Sea Atlas or Waterworld, by Johannes van Keulen (Amsterdam, 1682) (detail)

{National Maritime Museum, Greenwich, London}

Fig. 11 – ‘The Islands of Scilly’, from Great Britain’s Coasting Pilot, by Greenvile Collins (London, 1693). The lighthouse on St Agnis (St Agnes) was nine miles out of position (detail)

{National Maritime Museum, Greenwich, London}

Fig. 12 – The Indian Ocean, from The Great and Newly Enlarged Sea Atlas (Amsterdam, 1682), showing Europeans’ incomplete knowledge of the coastline of Hollandia Nova (Australia) (detail)

{National Maritime Museum, Greenwich, London}

Fig. 13 – Navigation instruments used in the late seventeenth century, from Practical Navigation, by John Seller (London, 1672) (detail)

{National Maritime Museum, Greenwich, London}

[an] abundance of grampuses, which are seldom far from land; but towards evening we had a more evident token, to wit, a little tired bird, something like a lark, came on board us, who certainly is an American, and ’tis likely was ashore this day.

The lead and line (Fig. 10 (#litres_trial_promo)) – a lead weight attached to a long rope that was dropped at regular intervals to check the depth and nature of the seabed – gave further help. North Sea sailors, for example, boasted that they could tell west from east from the pebbles that came up with the lead (which had a hollow base ‘armed’ with tallow to pick up seabed samples): those in the west could be broken between one’s teeth.

Lead, log and lookout worked well for coastal and short journeys, but might not be sufficient for longer ones. As European navigators embarked on increasingly ambitious voyages, often spending months in water too deep for sounding, they began to look to other methods for fixing their position. Being able to fix latitude and longitude with some degree of accuracy became more important.

One consequence of being unable to measure longitude directly was that seamen sensibly chose quite conservative routes. For example, if a ship set out on what the officers believed to be a direct course to its destination, there was the real danger that they would arrive at the correct latitude but find they had missed the destination. Unfortunately, they would not know whether they had sailed too far to the east or too far to the west, and so would not know which way to turn. The usual practice became to aim well to the east or west at the outset. Once the ship reached the latitude of their destination, they would ‘run down the latitude’ on a westerly or easterly heading, confident that landfall lay ahead. The buccaneer and explorer William Dampier (1651–1715) recorded using this method of latitude sailing on the Batchelor’s Delight in 1684:

we steered away N.W. by N. intending to run into the latitude of the Isles Gallapagos, and steer off West, because we did not know the certain distance, and therefore could not shape a direct Course to them. When we came within 40 minutes of the Equator we steer’d West ...

It was a longer journey but they arrived safely a couple of weeks later.

On some routes, latitude sailing was a matter of safety. Approaching the south-west coast of India from the Cape of Good Hope, for example, trading vessels needed to avoid the dangerous waters near the Maldives and the Laccadive Islands (Lakshadweep). The recommended course was to keep west to a latitude of 8° or 9° North, where there were safe channels running east to the Indian coast. Ironically, the predictability of latitude sailing made it dangerous in wartime, when enemy ships simply waited at the appropriate latitude for victims to sail to them, a tactic employed by French privateers off the Windward Islands of the Caribbean.

Mariners’ knowledge and skills, and the quality of their instruments, were crucial for effective navigation, as was the accuracy of charts and geographical data in printed manuals. However, these could be in error, even for areas close to home. Greenvile Collins’s 1693 chart of the Isles of Scilly from his Great Britain’s Coasting Pilot (Fig. 11 (#litres_trial_promo)), for example, placed the St Agnes lighthouse nine miles out of position, while the Philosophical Transactions, the Royal Society of London’s journal, warned in 1700 that the information normally issued for ships heading into the English Channel was dangerously misleading. In less familiar waters, charts were likely to be even more unreliable or incomplete: it would not be until the nineteenth century that Australia’s coastline would be fully drawn on European charts (Fig. 12 (#litres_trial_promo)).

Nonetheless, mariners had a set of methods that brought together centuries of accumulated seafaring knowledge with instruments and techniques that could be used to fix a ship’s position and course, and navigate it safely from A to B and back again (Fig. 13 (#litres_trial_promo)). The staple was dead reckoning, the only routine method of determining longitude until the end of the eighteenth century, and the dominant one long after that. It was straightforward, used a relatively inexpensive suite of instruments and worked well enough in most situations.

Early attempts to measure longitude

While most mariners could not determine their longitude at sea with the tools normally available, there were occasional attempts to do so, since the theories were sound. The most obvious approach was to use eclipses, which were predictable and simultaneously visible from different locations. By comparing the local time of the eclipse on a ship with the predicted time at a specific place, noted in astronomical tables such as Regiomontanus’ Ephemerides or Zacuto’s Almanach Perpetuum, a mariner could work out the longitude difference from that place.

Eclipses had long been used for observations on land, including an ambitious project of the 1570s and 1580s to fix the positions of different parts of the Spanish empire and improve the maps and charts held secretly by the Council of the Indies, the governing body for the Spanish colonies in America. The scheme relied on local officials building a simple moondial and marking the position of the Moon’s shadow on the dial when the eclipse began and when it ended. They then copied the marks onto paper and sent them with details of the length of the Sun’s shadow at noon back to Spain for analysis. It was perfect for keeping sensitive cartographic information secret but the data was fiendishly complex to process and was riddled with error. A more successful project was Philipp Eckebrecht’s world map of 1630, which used lunar eclipse data to plot many of the locations and was the first to equate one hour of time, astronomically determined, to 15° of longitude.

Fig. 14 – Two English ships wrecked in a storm on a rocky coast, by Willem van de Velde the Younger, c.1700

{National Maritime Museum, Greenwich, London}

Eclipses could not provide a routine solution at sea, however, since they occur infrequently, although they could be tried out occasionally. Christopher Columbus made observations twice in the Caribbean, in 1494 and again in 1503, although his results were not impressive in terms of accuracy. That said, his observations were more in the way of experiments and were taken at anchor, rather than as part of routine navigation at sea, for which he used dead reckoning. Nonetheless, he did believe that ‘with the perfecting of instruments and the equipment of vessels, those who are to traffic and trade with the discovered islands will have better knowledge’.

Alternatively, eclipse observations from a ship could be compared with observations taken at another location, but only when the results could be brought together at a later date. On 29 October 1631, a Welsh explorer, Thomas James, viewed a lunar eclipse from Charlton Island in what is now Nunavut, Canada, during a voyage in search of the North-West Passage. Meanwhile, the mathematician Henry Gellibrand observed it at Gresham College, London, and was later able to calculate the longitude difference from James’s figures as 79° 30' (a modern reckoning would place James’s position as 79° 45' west of Gresham College). Gellibrand considered this an impressive result that augured well for future advances in the art of navigation.

Almost forty years later, John Wood used eclipses of the Moon for on-the-spot longitude determinations when he was master’s mate on John Narbrough’s 1669–71 expedition to the Pacific, which was instructed to bring back geographical information and lay the foundations for future trade in South America. Observations at sea of a partial eclipse on 26 March 1670 gave the longitude of Cape Blanco (Cabo Blanco in southern Argentina) as 69° 16' W, while today’s value is 65° 45' W. Another observation a little further south on 18 September placed Port Desire (Puerto Deseado) at 73° W, the correct longitude being 65° 54' W. Wood also measured the position of the harbour of St Julian (Puerto San Julián, also in Argentina) from a conjunction of the Moon and Mars, calculating a longitude of 75° W, compared with a modern value of 67° 43' W. The observations showed significant errors by modern reckoning, not surprising given the instruments and data available, but they did demonstrate that determinations of longitude could be made while on expedition. While the infrequency of eclipses meant that they would never be routinely useful, other observations of the Moon had the potential to be used on a more regular basis and, as discussed in Chapter 2, some attempts to try them out were made in this early period.

Error and loss

Shipwrecks had many causes, just as they do today. Storms were a persistent problem but human error, including navigational mistakes, was also common. In many cases this was not simply about longitude determination but arose from a range of factors causing uncertainty as to a ship’s position and surroundings. Without proper charts, no amount of position fixing could prevent disaster.

Fig. 15 – Sir Cloudisly Shovel in the Association with the Eagle, Rumney and the Firebrand, Lost on the Rocks of Scilly, October 22, 1707

{National Maritime Museum, Greenwich, London}

Problems could easily arise in relatively unfamiliar parts of the world, and might be compounded by hostile weather and unknown currents. This was something that William Dampier, the first person to circumnavigate the world on three separate occasions, discovered repeatedly in a turbulent seafaring life. Dampier ventured into the Pacific for the second time in 1703 in command of the St George, as part of an ill-fated privateering party with the Cinque Ports. As the ships rounded Cape Horn, storms hit with their expected venom. The St George attempted to crawl its way around the Cape but its position was soon uncertain as the winds took it wherever they wished. What happened next depends on whose account one believes. According to William Funnell, an officer whom Dampier later accused of desertion, Dampier ordered the ship north once he believed they were to the west of Cape Horn but two days later it turned out, ‘contrary to all our expectations’, that they were still five leagues east of Tierra del Fuego.

Dampier saw it differently. While he conceded that there was some uncertainty about their east–west position, sighting Tierra del Fuego was not so unexpected:

for it is well known the Evening before, I told them we should see Land the next Morning, that of Terra del Fuego, the South Part of it: Now I look upon that to be a greater Mistake, to take one side of the Land for the other, than ’tis to be mistaken that we were Westward of the whole Island, and miss his Longitude ...

In any case, they were forced to brave the Horn once more.

Their troubles did not end there. Having separated from the Cinque Ports, the St George headed north towards the Juan Fernandez archipelago, off the coast of Chile, which was a regular rendezvous and watering spot for ships entering the Pacific. The typical approach to Juan Fernandez was to run down the latitude from the coast of Chile but, according to Funnell, the St George sailed right past because Dampier failed to recognize the islands. They finally returned after three days without sight of land, only to find the Cinque Ports safely anchored there. Dampier’s information and memory had led him astray. Incidentally, one of the sailors on board the Cinque Ports was Alexander Selkirk, the inspiration for Daniel Defoe’s Robinson Crusoe, who decided he would rather be abandoned alone on an uninhabited island in the archipelago than remain on the unseaworthy Cinque Ports. Despite the privations of life on the island, it proved to be wise decision as the Cinque Ports foundered later in its journey.

Mariners feared Cape Horn with reason, and the same was true of the western coast of Australia, which is littered with offshore reefs and islands that saw the demise of many ships plying their trade between Europe and Asia. The most notorious incident followed the loss of the VOC ship Batavia on its maiden voyage. Batavia sailed from the Netherlands in October 1628 and was in the southern Indian Ocean eight months later, heading along the recommended route eastwards before turning north for Java once it reached the correct longitude. By 4 June, it was approaching the Houtman Abrolhos, a known hazard off the west coast of Australia named ten years earlier by Frederik de Houtman, from the Portuguese abre os olhos, meaning ‘open your eyes’. The Batavia’s pilot knew he was approaching the reefs but seems to have ignored the danger signs and the ship struck. Of 322 on board, forty drowned during the shipwreck, and more than 110 men, women and children were killed as they awaited rescue in a tale of mutiny and murder that made for sensationalist reading back in Europe.