Navigation – Plan du site

European Navigation, Nautical Instructions and Charts of the Cochinchinese Coast (16th-19th Centuries)

Navigation, instructions et cartes nautiques européennes pour la côte de Cochinchine (xvie-xixe siècles)
Charlotte Minh Hà Pham
p. 101-129


La côte du Vietnam se trouve au cœur des routes maritimes qui relient la Chine et l’Asie du Sud-est. Durant les xvie-xixe siècles, alors que les Seigneurs Nguyễn développaient le centre du Vietnam, aka Cochinchine, le commerce maritime fleurit de concert avec l’arrivée des Européens dans la région. Les voyages entre l’Europe et les Indes orientales étaient longs, les navigateurs devaient maîtriser de nouvelles routes, les voyages devaient s’effectuer rapidement afin de faire face à la compétition mais surtout pour s’accorder au rythme des moussons qui ouvrait la porte à la mer de Chine du Sud et qui permettait d’atteindre la Chine. Les cargaisons étaient précieuses, la vie à bord difficile et les mers parcourues étaient inconnues. Ainsi, cartographier les mers était, afin de minimiser les risques, une besogne fondamentale pour les compagnies des Indes qui se faisaient face entre le xvie et le xixe siècle. L’article suivant examine les routes maritimes le long des côtes du Champa et de la Cochinchine à travers ces derniers siècles et surtout, les cartes nautiques et instructions nautiques que les capitaines et navigateurs européens possédaient afin de naviguer en sécurité dans les eaux périlleuses de la mer de Chine du Sud. Il existe très peu de sources sur la cartographie en Asie du Sud-Est, les sources en langues non européennes sont limitées et il en existe encore moins sur les cartes nautiques vietnamiennes en particulier. Les routes locales et les mouvements maritimes des pêcheurs, officiers de marine, ou marchands Chams et Cochinchinois ont été jusqu’à présent moins établis que ceux des Européens et moins facilement déchiffrables. Les marins européens se basaient pourtant sur les connaissances locales des marins malais, chinois et vietnamiens qui les guidaient dans les ports, pilotaient leurs bateaux et qui leur fournissaient parfois des conseils de navigation. Ainsi, la chronologie qui suit permet également de mettre en exergue le fait que le tracé de la côte vietnamienne sur des cartes nautiques n’est pas seulement le produit des efforts d’une seule nation européenne mais le résultat d’une conjonction globale qui prend ses sources dans l’expérience locale.

Haut de page

Texte intégral


  • 1 The terms “Champa” and “Cochinchina” refer to different geographical regions at different periods o (...)
  • 2 “Champa” refers to the extent of the former polity of Panduranga in the 16th and 17th centuries, fr (...)

1It was due to a storm in mid-September 1516 that the Portuguese made their first contact with Vietnam. The fleet, under the command of Fernao Perez de Andrade, was sailing off the coast of Cochinchina1 when a storm forced the captains of the vessels to seek shelter somewhere along the coast of Champa.2 Being caught in the transition period between the two monsoons, with the winds progressively shifting towards the South-West, de Andrade decided not to push further north to China but to return to Melaka (Barros 1628: 40v-41v).

  • 3 Chinese author Liu Xu, in his Old history of the Tang Dynasty, mentions that Chinese officials cont (...)

2At that time there was little European knowledge about the waters, climatological conditions, and routes of the South China Sea. Early European seafarers travelled on board local ships or were guided by local pilots as they started to draw up their own maps and compile their own sets of nautical instructions for the area. Drawing on the local knowledge and experience, Europeans followed the traditional route skirting the coast of Vietnam,3 which was considered as the safest passage. As they sailed more regularly to the East Indies, the South China Sea coastline became progressively better represented in maps, charts and rutters, and Cochinchina and Champa were revealed to the world as locations were progressively marked down and the number of identifiable landmarks increased over time. The crucial Vietnamese landmarks of Pulo Condore, Cape Padaran, Cape Varella and Pulo Canton featured on most maps and rutters of the region since the 16th century. As the centuries passed, the nautical instructions gradually became more precise, while the frequentation of this main maritime route did not change much.

  • 4 About cartography in Southeast Asia, see Fell (1991), Schwartzberg (1994) and Suarez (1999). About (...)

3When they arrived in yet uncharted waters, Europeans had to familiarize themselves with the nearby coast and sailing conditions, and for this, relied on the local knowledge of Malay, Chinese and Vietnamese seafarers to build up their confidence in sailing their new routes and itineraries. Yet, the local routes and maritime movements of Cham and Cochinchinese fishers, naval officers and traders, are so far, less clearly delineated than those of European navigators and not readily decipherable. There is little record of Vietnamese seafaring, of their maritime routes, or of their nautical aids and cartography. There are only a few sources in non-European languages on navigation in the South China Sea and even less on the charting of Vietnamese waters.4 In European languages, on the other hand, there are abundant sources from the 16th to the 19th centuries, including ships’ logs, travellers’ accounts, mission reports and nautical instructions (Pham 2015). When drawing upon Portuguese, Spanish, Dutch, British and French documents together, a continuity in the development of the European knowledge of the South China Sea over the course of three centuries appears. It reveals how Dutch maps were based on earlier Portuguese and Spanish material, and how French and British hydrographers collaborated to produce the most complete set of nautical instructions. It puts forward a longue durée of navigation on the South China Sea and steps away from a nation-centred narrative. Bringing together the charting of this region by European nations highlights the fact that charting the Vietnamese coastline was not the effort of any one particular European nation but rather the result of a global conjunction that had also roots in local experience. It is a good starting point to tease out a form of common knowledge of the seas.

4This paper introduces the main maritime route along the coasts of Champa and Cochinchina between the 16th and 19th centuries, and discusses the nautical knowledge, charts and instructions that European navigators possessed in order to navigate the perilous waters of the South China Sea. It also provides an opportunity to recognize the hitherto invisible experience of the local seafarers because their traditional knowledge and experience formed the basis upon which European navigation grew. Concepts of geography and navigation, tools and instruments (or lack thereof), and seafaring imperatives differed between European and local sailors, but the seas they sailed upon were the same. So in the absence of studies on Cochinchinese navigation, European charts and records can throw light on the nature of this coast, on the main traditional routes, and act as a point of entry to ultimately approach the traditional practices of Cochinchinese seafarers.

The Early Routes

5The Portuguese seafarers of the 15th and 16th centuries explored the earth’s oceans and opened maritime routes to the Americas, Africa and the Orient. In 1511 they settled the Straits of Melaka, and from there, they sailed along the coast of Champa and Cochinchina to reach Macau, their future commercial hub.

  • 5 In the 16th century, the navigational technology did not allow navigational precision greater than (...)
  • 6 See Manguin (1972: 74, 109).
  • 7 As seen in twenty-four rutters for the region, in the series entitled Advertencias para a navegacae (...)

6Because the early Portuguese rutters were not detailed enough to allow captains to navigate and sail their ships in uncharted waters (Manguin 1972: 171), because the charts at that time were inexact and drawn on a very small scale (< 1: 50 000 000), and because in the early 16th century navigation instruments were not at all accurate,5 it was common practice to board local ships or hire local pilots.6 Recurrent trips gave Portuguese sailors unprecedented seafaring expertise in the East Indies, but during the early stages of the Portuguese expansion, their activities nonetheless relied primarily on local knowledge, essentially Chinese maps and instructions provided by local Malay pilots. The earliest known European rutter of a journey from Melaka to China is dated between 1511 and 1513 and was written by Portuguese Francisco Rodrigues, based on his own observations as well as on indigenous sources of information (ibid.: 52-53). As they more regularly frequented the region and followed the traditional routes used for centuries by regional traders, more rutters and charts were produced. The precision and clarity of the instructions increased as more landmarks were identified and combined with information on depths and latitudes, and that measuring distances and positions became a little more accurate over time (Manguin 1972: 157).7

7Portuguese traders and navigators sailed along the pre-existing local routes and networks of ports and stopping points supported by regional trade and did not push for faster and more direct routes. The safest and most common route that joined Melaka to Macau involved hugging the coast of Cochinchina for its entire length in order to avoid the Paracels and other unknown hazards of the South China Sea. This traditional route did not cross the South China Sea. The instructions were initially composed of just a few conspicuous markers; from Melaka, once past the Straits, the instructions were to sail for about a week until sighting the islands of Pulo Condore [Côn Sơn]. From there, pilots were to navigate north-east, keeping a depth of 20 fathoms until sighting the island of Pulo Cecir do Mar [Phú Quý]. The ships were then to pass between the island and the coast of Champa. Approaching closer to the coast and rounding the conspicuous Cape Padaran [Mũi Dinh], ships would follow the coast until reaching the very high and visible Cape Varella [Mũi Đại Lãnh]. From there, following the coast at a distance of about five to ten miles for a few days, the instructions were to pass by the island of Pulo Cambir [Cù Lao Xanh] to finally reach the island of Pulo Canton [Lý Sơn]. In the 16th century, it was from there that ships crossed the sea to reach the island of Hainan and then Macau. The voyage between Melaka and Macau lasted between 30 and 40 days. The main hazards along the route, the Mattheus Brito shoal, the reefs surrounding Pulo Cecir do Mar, the rocks and islets in the bays of Nha Trang, and the great extensive Paracels shoals, were also noted in these early instructions. As local sailors had done for centuries when sailing to and from China, the Portuguese hugged the coast of Vietnam following an easily recognizable chaplet of islands and conspicuous high capes.

8Over the years, although the route remained the same, the collections of rutters passed from hand to hand, and included more information as each pilot who sailed in the area added new details, greater clarity and precision and new toponyms to the preceding ones (Manguin 1989).

  • 8 These toponyms can be related to the Malay words barahla, pulau, sisir, kambing, etc. (Manguin 1972 (...)

9The European breakthrough and expansion in the region in the 16th century definitely owes a great deal to the Portuguese merchants, whose legacy is still present in some of the toponyms in Vietnam. For example, the Cape Vũng Tàu from a distance looks as if formed by five peaks and was therefore called Cinco Chagas, referencing the five wounds of Christ (Manguin 1972: 93, note 5). The Portuguese name became Sinkel-Jacques in Dutch, Cape Saint-Jacques in French, and this in turn became Cape Saint-James in English. On the other hand, many toponyms that stem from the local Malay/Cham names have survived, thanks to their Portuguese rendering, such as Cape Varella, Cape Padaran, Pulo Cecir, Pulo Cambir.8

The Mythical Danger of the Paracels

  • 9 The Scorpion’s Tail was also an imaginary bank, believed to be connected to the Paracels; “It is fr (...)

10The name Paracel also comes from the Portuguese: o parcel meaning “the reef” (Manguin 1976: 114). It was believed to be a single feature that extended all the way from latitude 17°N to latitude 11°N, stretching from the south of Hainan Island and almost reaching down to the proximity of Pulo Cecir do Mar. There, the few rocks and breakers in the vicinity were seen as joining the islands Pulo Cecir do Mar and Pulo Sapata in the shape of a scorpion’s tail or Rabo de Lacra.9

  • 10 See Manguin (1976) who briefly considers the question, as well as some famous Chinese rutters that (...)

11The traditional route hugging the coast was due in no small measure to the great fear that the navigators had of being lost or losing a ship in the Paracels. This anxiety had been perpetuated since the first journeys of the Europeans in the region, and was probably passed on down from Malay, Chinese or Cochinchinese pilots to Portuguese navigators, then confirmed by their own experiences of hardship and shipwreck within the reefs and shoals.10As European seafarers made their way into the region, and copied, handed down or stole rutters and navigation instructions from one another, this large imaginary rectangle of mythical dangers persisted.

  • 11 Note that if the Paracels were not surveyed and charted until the 19th century, it is also because (...)

12The deep ignorance about the size and precise position of the Paracel and Spratly islands made them a terrible hazard that cost the lives of sailors and caused the loss of ships and valuable cargos, and that was better avoided. On top of being uncharted, the main danger was the fact that the reefs and islands emerge in the surf, are barely visible in cloudy weather or at night, and that most of them fall abruptly to great depths, so that no soundings announce their presence. Furthermore, heavy seas, storms and even typhoons also made navigation across the South China Sea very dangerous, despite local knowledge of the weather conditions, and despite the first theoretical understanding of the general monsoon system by Edmund Halley (1686: 153-168). Until at least the 18th century, there was very little scientific understanding of the meteorology of the South China Sea, and this combined with the uncharted hazards, were the basic reasons why in the collective nautical knowledge of the South China Sea, it was safer to sail along the coast of Vietnam and why it remained the principal route until the 19th century.11

In the 17th Century

  • 12 In the 15th century, King Joao II of Portugal tried to enforce a “policy of secrecy” (sigilo) to ke (...)
  • 13 See the rutters of the mid-17th century in the Advertencias (Manguin 1972: 91-103; 107-120; 122-125 (...)

13Throughout the 16th and early 17th centuries, having gained experience through recurrent explorations of the Atlantic and Africa, the Portuguese were the most accomplished and expert seamen, and their experience and knowledge of navigation was well sought after.12 With “the help of God” and the concise (and imprecise) instructions borrowed from locals, the Portuguese navigators progressively became quite familiar with the waters bordering Champa and Cochinchina. If the earliest Portuguese rutter of Francisco Rodrigues comprised just eight markers for the route between Melaka and Hainan, the mid-17th century rutters of Francisco Pires comprised triple that amount of markers and descriptions. The itineraries continued to increase in information, slowly becoming slightly more furnished and descriptive, with new landmarks noted down (i.e. Cinco Chagas, Rao Rao, Sidrao, Sinco Irmas, Isla de Caranguejo, Comori), with visual portrayals and nuanced descriptions of coastal locations and navigational hazards, as well as with depths and distances, and personal comments about local conditions and tides.13 Although not being precise navigation compared to later 19th century standards, these earlier rutters were starting to improve progressively from what had been available up until then.

14By the early 17th century the Portuguese were not the only Europeans navigating along the coast of Champa and Cochinchina anymore. The new century witnessed the arrival of both British and Dutch envoys and their initial attempts to undertake trade negotiations with the Nguyễn Lords in Cochinchina and the Trịnh Lords in Tonkin. Spanish, Dutch, British and French ships, all made their way via the traditional route, as the Portuguese did.

  • 14 The rutters in chapters 20, 24 and 44 of the Reys-gheschrift are copies of some Portuguese rutters (...)
  • 15 Lucas Janszoon Waghenaer’s Spieghel der Zeevaerdt (1584/1585) and Het Licht der Zee-vaert of Willem (...)

15In the early days of Dutch navigation in the East Indies, the Dutch instructions—secretly provided to the pilots—were still mostly copied from Hispanic sources. The earliest Dutch set of nautical instructions for Southeast Asia dates to the end of the 16th century. It was compiled by Jan Huygen van Linschoten, Dutch merchant, traveller, and secretary of the Archbishop of Goa, between 1583 and 1589. His Reys-gheschrift (Linschoten 1595) includes rutters for the region, all copied from Portuguese material14 and was soon translated into English and French. In the 17th century, the early input of the Portuguese was nonetheless superseded by the Dutch cartographic tradition which, following the great Mercator and Ortelius, reached its zenith. Gradually, the Dutch changed the nature of the navigation aids by creating nautical charts per se in which various information such as coastlines, islands, harbours, as well as water depths and hazards were graphically compiled (Zandvliet 1994).15

  • 16 For example, Linschoten obtained information about the navigation between China and Japan from his (...)
  • 17 Dutch navigation instructions were not in wide circulation, in contrary to the Portuguese instructi (...)

16Information for charting the East Indies was available through the Vereenigde Oost-Indische Compagnie (VOC) nautical expeditions.16 Famous watercolourist and chart maker Johannes Vingboons produced, between 1640 and 1667, maps, coastal profiles, charts and plans based on the material of the VOC returning crew members. His manuscripts include one of the first charts centred on Cochinchina which also provides navigation information along the coast of the country that had not been presented before in such a manner (for example, depths are provided for every 5 to 10 nautical miles (nm), river mouths are indicated, and zones of shallow water are depicted, the main islands and capes are named). Some of Vingboon’s work was included in Joan Blaeu’s Atlas Maior but many of their respective productions remained quite exclusive and only benefited VOC mariners.17

  • 18 Asian ships followed the same route, however tended to veer to Hainan from Pulo Canton, as attested (...)
  • 19 An estimated 247 VOC ships were lost in the 17th and 18th century in the East Indies (Hol et al. 20 (...)

17To link Batavia to Fort Zeelandia (Taiwan), Hirado or Deshina in Japan, ships still sailed near Cochinchina, passing Pulo Candoor, Baij Pandoran, Houck van Warella, Campello [Cù Lao Chàm], Baij van Thoron [vịnh Đà Nẵng] and the trading town of Phaijfo [Hội An]. To reach Hainan, the Europeans tended to head for the islands of Champello instead of Pulo Canton before continuing their routes to China and Japan.18 In 1622, the fleet under Admiral Reijersen left Baij Comorin [vịnh Cam Ranh] in June 1622 and reached Macao in seventeen days. The Haarlem and the Leyde (1601), the Brouwershaven (1633), the Amsterdam (1637) and the Delfshaven (1651) also covered the coast of Vietnam in about a week. Not all ships were so lucky, and at least a dozen VOC ships were lost19 on journeys in the South China Sea, including in the Paracels, which were still depicted as a long rectangle on charts and world maps.

18At that time, to determine position at sea was still only relative, bearings and distances were at best approximate and land outlines were only roughly sketched. Navigation was not yet accurate and precise (bearings to within 5 degrees or so, distances to within plus or minus 10 miles). In general, 17th century ships on very long journeys needed to stop, and therefore had great interest in places for breaming, careening and repairing, reprovisioning, recuperating the sick and watering etc., but on shorter journeys, or if they had done so in Melaka or Pulo Condore, to sail offshore was not a problem. In fact, captains preferred if possible to sail at a certain distance from the coast (particularly when uncharted) and sought to approach only in case of emergency or in fear of storm or typhoon. Therefore, the Paracels were still the main obstacle to make the end run to and from China without stopping, and were the main reason why European ships still sailed closer to the Cochinchina coast.

Crossing the South China Sea

19In the 17th century, although the Paracels were still deemed a mysterious expansive hazard stretching from north to south, there were instances when navigators nonetheless guided their ships across the South China Sea, passing through this imaginary rectangle. From Pulo Sapata, they sailed south of the “Scorpion’s Tail” to go to Manila and vice-versa, as did the famous adventurer and buccaneer William Dampier, on board the privateer Cygnet commanded by Captain Read. They left Luconia [Manila] on 26 February 1687, during the NE monsoon, and boldly crossed the sea.

[…] in latitude 14 degrees north […] we began to steer away for Pulo Condore, and we steered south by west […]. In our way thither we went pretty near the shoals of Pracel [sic] and other shoals which are very dangerous. We were very much afraid but escaped them without so much as seeing them. (Dampier 1697: 389)

20Dampier and his shipmates reached Pulo Condore on 13 March 1687 after a 15 days crossing. Three months later in June, intending to go back to Manila they attempted to cross the sea again, but the winds pushed them towards China instead. They managed to reach the Pratas Island, miraculously avoiding the terrible Pracel [sic] (Dampier 1697: 403).

  • 20 In the 1660s, four junks per year sailed between Manila and Cochinchina (Li 1998: 76). Several inst (...)

21Crossing from the coast of Cochinchina to Manila is also attested in the diary of the Franciscan friar Jesus de Escalona who, in June 1640, left the port of Turon to go back to his convent in Manila (Perez 1916). Trade between Manila and Cochinchina had reached its peak in the 1660s and such crossings became common at the end of the 17th century.20Italian Jesuit Gemelli Careri noted in 1695 that from Pulo Condore “all ships that go to Manila are used to recognize this island” (Cadière 1930: 305). Less fear was associated with the Paracels by then. Through trial and error, knowledge grew about areas of relatively open sea, sailing through using Pulo Condore and from thence, Pulo Sapata as a departure or arrival point.

Sellers’ English Pilot

22When William Dampier and Captain Read crossed the South China Sea, apart from stolen charts and sailing directions from captured Spanish vessels (Williams 1997: 107), they could have been in possession of John Seller’s English Pilot (Book III on the Orient), which was the only available set of nautical instructions for the East Indies in English made publicly available, or of his Atlas Maritimus, which was a compilation of nautical charts. Both were published for the first time in 1675.

23John Seller (1632-1697), British mathematician, compass maker and map seller, compiled over the years a six-volume set of sea charts. However, Sellers’ charts were counterfeited from old Dutch copper plates (Verner 1978: 139), the map of the East Indies included in his Atlas Maritimus was essentially copied from a 1666 map by the Dutchman Goos (Suarez 1999: 209). After John Seller died, his associates John Thornton Senior and Junior pursued the publication of the various volumes of the English Pilot. Thornton’s new edition was brought to fruition in 1703 and superseded Sellers’ version. Thornton’s version included not only charts, tables of magnetic variation, of longitude and latitude, but also some extensive sailing directions and numerous coastal profiles. It was also a forerunner to a “merchant pilot’s guide”, including the occasional bit of advice on how to conduct trade (ibid.: 210). This made the English Pilot a most essential guide. It was so well received that it remained the standard reference for navigation instructions and was published in eleven subsequent editions throughout the 18th century.

24At the end of the 17th century, apart from Sellers outdated charts and instructions, the tools to support navigation in Southeast Asian waters were still very limited (or unavailable in the case of the Dutch material). Geographical knowledge of Cochinchina was nonetheless growing, thanks mainly to accounts of travellers and missionaries. In 1679, one of the very first maps centred on the Kingdom of Cochinchina appeared in Jean Baptiste Tavernier’s Recueil (1679). This map despite being geographically distorted (particularly the region of the Mekong), comprising a number of inaccuracies, and being totally useless for navigation, showed Cochinchina in good detail.

The Missionaries’ Scientific Contributions

25Another map famous for framing the entire country, and remarkable for also depicting the coast and near-by islands as much as the land, was made by Jesuit Father Alexandre de Rhodes, who proselytized in Cochinchina and Tonkin between 1625 and 1645 (Rhodes 1650).

26Captains, naval officers, and sailors were not the only ones to sail Asian waters and to possess an acute understanding of the Cochinchinese coast. As European missionaries set foot in the country during the first half of the 17th century, their knowledge of the new land of their apostolate contributed greatly to the exploration and understanding of the area. Versed in faith as much as in science, mathematics, and cartography, missionaries were skilled in producing maps.

27Sometimes, missionaries knew the coast almost better than sailors. For example, in 1686, Abbot de Choisy on a mission to Siam with the embassy of Chevalier Chaumont, explicitly noted that the navigators on board relied on the maps drawn by father Bénigne Vachet who had travelled extensively in the area since the early 1660s.

M. Thevenot gave very beautiful nautical charts to the Jesuits; he had them copied on those from the King’s Library. Our pilots thought they were good, but were down on their knees when seeing the ones of M. Vachet. Everyone is after them to copy them. (Choisy 1687: 25-26)

28Father Bénigne Vachet knew the coast of Cochinchina extremely well. Indeed, between 1671 and 1686, Vachet had made the journey a number of times between the seminary of the Missions étrangères de Paris (MEP) in Ayutthaya and Nha Trang or Faifo on board local boats. Due to the waves of persecutions against Christians, and the necessity of having official passports to enter and leave the country, as well as the tight control exercised in the big ports of Cochinchina, missionaries like him needed to enter the country discreetly, avoiding big towns and bustling harbours. This meant they did not always go to the most common ports and on the contrary, were familiar with the secondary ports and small villages. By the end of the 17th century, thanks to the activities of the MEP and the Jesuits, Cochinchina became well known to the world and progressively more accurately represented on maps too.

  • 21 It was still based on the idea of simultaneous observation of a celestial event from two different (...)

29The Embassy of Chaumont (1685) was equipped with astronomical instruments to try to determine, on behalf of the Royal Academy of Paris, the position of Siam and fix the longitude of Lop Buri. This was an important endeavour because in these early days of navigation to determine a ship’s position in the open sea and therefore its longitude was a major problem, and successful sailing in Asian waters still relied on Providence. Latitude is relatively easy to determine at sea (to a certain level of accuracy, about a quarter of a degree, which is equal to plus or minus 15 miles) and became common practice in the 15th century. To determine longitude however was more difficult because longitude calculation depends on time and the main challenge was precisely to pinpoint time on a moving ship. While the missionaries methods were still old fashioned21 (and it was only over a century away before the problem began to be solved effectively), the Jesuits exercises and results for determining the longitude of places they sailed past were an important contribution to the advance of navigation science.

In the 18th Century: British and French Advances

30As the Dutch cartographic tradition slowly started waning at the end of the 17th century, in concert with less control by the VOC over the spice trade, the following century saw the French contribution increase through recurrent missions to Southeast Asia, while the British productivity and superiority in the development of the sea-atlases and nautical instructions took centre stage. Technically, it is during the 18th century that the challenge to determine longitude was taken on in England with the set up in 1714 of the Board of Longitude (Johnson 1989).

D’Après de Mannevillette and Dalrymple

  • 22 D’Après de Mannevillette was navy officer until 1759 and then appointed chief of the Dépot des cart (...)

31As commercial interests and competition reached fever pitch between the European East Indies Companies, the 18th century marked the beginning of the systematic publication of navigation instructions for recurrent voyages, and the systematic collection of shiplogs (Filliozat 2003: 6). Despite the global rivalry between the British and French empires, the nautical advances of their respective captains, merchants, and hydrographic services jointly contributed to the development of safe passages and routes around the South China Sea. The French contributed greatly to the improvement of navigation in the region, taking the lead with the establishment of the Dépôt des cartes et plans de la Marine in 1720. In 1745, Jean-Baptiste-Nicolas-Denis d’Après de Mannevillette, French hydrographer for the Compagnie des Indes,22 compiled the most up to date instruction manual for the eastern seas, the Neptune Oriental. He combined his own experience and observations with the content of the ship logs of the French Compagnie des Indes vessels, as well as with the information contained in numerous older rutters and ship logs he gathered from other seafaring nations, including the Dutch, Spanish, Portuguese, particularly Manoel Pimentel’s fundamental Arte de Navegar (1712), and also from Thornton’s English Pilot (Après de Mannevillette 1775: 430). The work of d’Après de Mannevillette was also ground breaking because he was one of the first navigators to use an octant, a new instrument that helped to calculate position and time, and because he systematically provided longitudes (as precise as they could be at that time) for all the positions he described, based on lunar measurements.

  • 23 He produced another map, dated to 1755-1759, but published in 1787. It represents the same portion (...)
  • 24 M. Didier was an engineer of the King, and commander on the island in 1721 to 1722, when the French (...)

32For navigation along the Vietnamese coast, it was by far the most complete and accurate set of instructions ever produced to date. It was also the first time that nautical instructions included maps of portions of Cochinchina which were produced to a larger scale (1: 350 000). The first one was made by René Julien le Floch de la Carrière and outlines the coast between Faifo and the island of Champello, and Hue, the Nguyễn capital.23 It provides soundings and rhumb lines, shows locations of capes and inner bays, and also includes profiles of the entrance to the Bay of Turon and Champello. On the same plate, d’Après de Mannevillette also included a chart of the island Pulo Condore (1: 70 000) with relief, water depths and a coastal profile, drawn by M. Didier between 1721 and 1722.24 At the time no other more detailed charts of the coast of Vietnam were available, and the country was still only depicted in larger maps of the East Indies, and represented in very small scale with few details. These advances and maps were still not sufficient to contribute to an entirely accurate navigation.

33In terms of nautical instructions, d’Après de Mannevillette’s work became the standard reference (Filliozat 2003) for navigation beyond the Cape of Good Hope. It was translated and annotated by the British William Herbert in 1758 under the title New Directory for the East Indies (Herbert 1758). At that time, the most commonly used set of charts for British navigators was the East India Pilot (Sayer & Benett 1778), edited by Robert Sayer for the first time in 1750, but no instructions accompanied the charts. It did not contain a good chart of Cochinchina, only a copy of the chart of le Floch de la Carrière. Dutch charts certainly existed too, but were still not publicly available. In 1753, Johannes II Van Keulen, the appointed official chartmaker of the VOC, completed a work his grandfather had started in 1681; a collection of maps, the Zee Atlas, and nautical instructions, the Zee Fakkel. Johannes II completed the 6th volume of the sea pilot, using material from the secret files of the VOC. It was however only published in the “secret atlas of the VOC for their own internal use” (Zandvliet 1994: 1461-1463).

34Over the course of several voyages between 1752 and 1765, Alexander Dalrymple, on board the Cuddalore, as a passenger initially, and then commander (1761 to 1763), before being appointed hydrographer of the British East India Company, sailed the South China Sea, and compiled enough nautical information to create a new and more detailed chart of the area. Dalrymple’s chart, published in 1771 within his Memoir of a Chart of the China Sea was the best and the first to accurately map the contours of the South China Sea (c. 1: 4 500 000). It stemmed from d’Après de Mannevillette’s work, of which he collated various sources of information. When possible, he eliminated errors, and added tracks of ships whose data he trusted to show where it was safe to go. Unfortunately, Dalrymple “traced the Coast of Hainan and Cochin-China very faintly […] not having hitherto laid down these coasts from my own observations” (Dalrymple 1786: 9). On the other hand, he included for the first time the main islands of the Paracels in relatively good detail (better than previously available). It was drawn thanks to a plan made by a Cochinchinese pilot that Dalrymple obtained in Manila.

35When in 1775 d’Après de Mannevillette published a second and revised edition of the Neptune Oriental, he included a copy of Dalrymple’s chart and memoir. Both hydrographers had great respect for one another and each praised the other in their respective works. It is mainly thanks to the contribution of both Dalrymple and d’Après de Mannevillette, who sifted through all the ships’ logbooks, that sailors could now sail with more confidence in the East Indies. Yet, despite their combined efforts, by 1775, the coast of Cochinchina was still charted with many inaccuracies. During his travels, Dalrymple did not stop along the coast of Cochinchina, and d’Après de Mannevillette, despite five campaigns to India and China (Filliozat 2003: 7), was unfamiliar with the coast:

I must admit that we only have uncertain knowledge, and this is what needs particular attention from the ones who will skirt this coast, or from the ones who would want to land there. (Après de Mannevillette 1775: lvi-lvii)

  • 25 Huddart’s version, while repeating the content of the Neptune Oriental, re-organised the layout, th (...)
  • 26 They all repeated the same mistakes. For example, the inexistent island “Pulo Cambir de Mer”, which (...)

36In 1780, the second edition of the Neptune Oriental was in turn methodised, corrected and further enlarged by British Samuel Dunn, based on the previous edition of William Herberts’ work, to form the 5th edition of the New Directory. Dunn added important sections about the effects of the monsoon, about the currents, and other climatic factors as well as a few more details about the routes that were not present in the Neptune Oriental. It was then superseded by Joseph Huddart’s Oriental Navigator in 1794.25 Because the weather was not totally understood and systematized yet, the information provided only intended to explain the main patterns and the normal range of weather variation. Sections of log books and narratives were included in the instructions to convey meteorological complexity and variability, and provide examples of how a voyage could be conducted at a certain time. For sailing along the coast of Cochinchina, the descriptions remained based, if not literally copied, from the Neptune Oriental.26 However, since vessel quality (ie: copper sheathing to protect the ship’s hull), provisions (improved methods for food storage) and crew management, and navigational knowledge had progressed since the 17th century, for the 18th century deepsea captains, there was less necessity to know the coast they sailed along in detail. The conspicuous landmarks and main hazards were known, and unless sailing to Cochinchina with a specific purpose (trade, diplomacy, religion), the information now available was sufficient in case something went wrong (crisis, ill weather etc.) to seek shelter and recognise the coast. Only when they stopped along the coast for business or due to ill weather, did captains had more soundings and distances measurements taken, harbours sketched, and the position of hazards checked, in order to complete the information already available in published instructions and charts. This is why the instructions for sailing along the coast of Cochinchina were mostly repeated without much improvement, and were upgraded only slowly.

Navigation on the Coast of Cochinchina

37For the late 18th century captains who actually navigated the coasts of Cochinchina, the instructions were therefore still not reliable enough. Guy-Pierre de Coenempren Chevalier de Kersaint, Captain of the frigate Dryade in 1788, did not trust the maps he had on board, judging they were “produced without much regularity” (Bizalion 1996: 133), particularly for the region near Cape Varella. Likewise, British Captain Blomfield, sailing the Indiaman Admiral Pocock in 1764, using one of the first editions Herbert’s New Directory noted that “[I] cannot find he carries you with any certainty further than Pulo Canton” (Lamb 1961: 80). Indeed, in the first edition of the Neptune Oriental and hence, in the translated version by Herbert, there is very little information about the navigation north of Pulo Canton, to Faifo and further to Hue.

We have no certain knowledge of this port [Turon], though it is the most trading one on the coast of Quinam, because Chinese, Tonquinese and the neighbouring natives are the only people trading thither. […] The descriptions I have met with of these coasts are reduced to some few instructions not particular enough to be of any use [...]. (Herbert 1758: 99-100)

38By 1793, Sir George Staunton, secretary of British ambassador Earl Macartney, on an embassy to China on board the Hindostan noted that even if a pilot was not necessary anymore to sail from Cochinchina to the Chinese coast, “few accurate accounts had hitherto been published of the navigation” (Staunton 1797: 383-384). Despite the Bay of Turon and the harbour of Faifo being the main points of foreign trade, where numerous missions, convoys and ships sailing past this coast would stop, the instructions to Vietnam at the end of the 18th century were still limited.

New Passages in the South China Sea

  • 27 The Macclesfield Bank is a easily recognisable, very large (6 448 km2) and submerged bank (shallowe (...)
  • 28 In subsequent navigation manuals such as James Horsburgh’s India Directory, this “east of the Parac (...)

39To sail to China by passing along the coast of Vietnam was still the main route traditionally followed. However, by the 18th century, a new return passage in the NE monsoon passing east of the Paracels, and therefore away from the coast, was now indicated. Although the exact extent and a full detailed composition of all the Paracel Islands was still unclear, the area was no longer considered as dangerous and as extensive as it was in the past. This was thanks first to Manoel Pimentel (Pimentel 1712: 413), and then to d’Après de Mannevillette who himself sailed as early as 1738 from the Grande Ladrone to Pulo Sapata. He highlighted the Banc des Anglais or Macclesfield Bank27 to the east, and demonstrated that the many islands and hazards believed to lie in the southern end of the Paracels were non-existent (1775: 460-461). This led him to suggest what he called an “eastern passage”28 that did not hug the coast of Cochinchina anymore but rather sailed a long distance from it, from Macao, via the Macclesfield Bank, across the sea, south of the Paracels, heading boldly towards Pulo Sapata and then Pulo Condore.

  • 29 Pirates in Southeast Asian waters have always been a problem; until the late 17th and early 18th ce (...)

40Sailing in deep water and off shore was always preferable for western captains on long-distance journeys, because it avoided all the hazards (known and unknown) which are usually found within 10 miles of a coast (Davies 2015, pers. Comm.). Secondly, the fear of pirates was also a good reason for keeping away from coastal areas, particularly in the 18th and 19th centuries.29 This passage however, was advisable only for “good and fast ships” sailing well into the NE monsoon, because when taking this “east of the Paracels” passage, the advantages of keeping sight of land, and of being able to reach safe haven in case of storm, heavy seas, or problem, were lost. In fact, although it appeared faster and more direct to aim for the Macclesfield Bank before bearing away for Pulo Condore, ships still preferred to follow the traditional route. The new route was actually as long, and most of the year, the weather conditions were very variable—apart from the peak of the SW monsoon when the route could be more expeditious. Therefore, this passage was not always recommended, particularly when bound from China and fully ladden.

41In the 18th century, since it was also introduced in the Neptune Oriental (Après de Mannevillette 1775: 539-546), the passage to Manila from Pulo Condore—as proceeded along by Dampier at the end of the previous century—had become common. Between 1749 and 1757 Pierre Poivre travelled regularly between Melaka, Canton, and Manila, passing along the coast of Cochinchina or across the sea from Pulo Condore to Manila. This route was also followed by the Dryade for example, which sailed in September 1788 directly from Pulo Condore to Manila. But if one did not sail far enough north of Pulo Sapata, to latitude 12°30'N, in order to then cross over to reach Cabra Island on the other side of the sea, the myriad of yet uncharted Spratly reefs and islets could prove fatal. This lead to the epic misadventure of the sailors of the Earl Temple, a British East Indiaman which wrecked in the area; she left Batavia on 16 May 1763 and less than a month later, on 7 June 1763, wrecked on the reefs of the Spratlys. A few crew managed to survive on an island with Cochinchinese fishermen who had also lost their ship, where they remained stranded together for a few months before sailing to the coast on a makeshift raft.

42When the Admiral Pocock, on her way to Canton in December 1764, unintentionally ended up in the Bay of Turon, she remained there for five months waiting for the onset of the SW monsoon to pursue her trip to China. While there, Captain Blomfield and Captain Arthur Gore took the opportunity to survey the Bay of Turon and to produce the first large scale map of the bay (1: 100 000). It was published in 1774 by the EIC hydrographic service, under the direction of Dalrymple. Along with the map of le Floch de la Carrière, it was the only map of the area available at the time. This map was unfortunately “merely a rough sketch […] [that] gives no information or instruction how to enter into the bay, and was […] found to be erroneous in many respects” (Staunton 1797: 325). This is why, when the fleet of Macartney stopped in Turon, John Barrow and Captain Parrish took the time to survey, as systematically as possible, and map the Bay of Turon and the islands of Champello. For almost a hundred years—until the advent of the French Protectorate and its hydrographic service—their map remained the most useful and detailed (ca. 1: 150 000-160 500).

43In the second half of the 18th century, further surveys were conducted to chart the region, including the survey conducted by the French naval officer Theobald René de Kergariou-Locmaria in 1788, which was crucial in determining the extent of the Paracels. From 1788 to 1789, Chevalier de Kersaint and Commander Rosily-Mesros, also engaged in hydrographic surveys and mapping some areas of southern Vietnam (Sài Gòn River, Cần Giờ harbour) but their works were not substantial, nor widely distributed due to the political unrest of the French Revolution.

  • 30 The English Pilot has about 30 landmarks whilst the Neptune Oriental has about 40.

44Step by step, edition after edition, the British and French worked concurrently to improve the conditions of navigation in the East Indies. Compared to the sets of instructions from the late 17th century, the quality of the instructions of the 18th century reached a higher degree of precision and quality. Navigation experience evolved from being mostly empirical and approximate to a qualitative exercise that produced detailed charts and instructions. The routes to the East Indies were now far more detailed with actual visual descriptions, with many more landmarks,30 and names for most of the bays, capes and islands. Maps were consequently also more precise and elaborate, and the quality of charts increased as more European ships sailed the region. As a new era commenced in terms of navigation technology, further knowledge was acquired in regard to the Paracels and hence, in routes to China for deepsea European captains who preferred to sail off the coasts, directly and faster to their destinations, motivated by major commercial interests.

In the 19th Century

Horsburgh and Dayot

45In the early 19th century, the British hydrographers were in the vanguard, replacing the superiority of the French, disrupted by the French revolution. Between 1809 and 1811, James Horsburgh published his authoritative work The India Directory. It became the main reference for 19th century navigation in Asian waters, and was re-edited a number of times until 1855. The India Directory is marked out by a quantum shift in detail, quality and organization. What was revolutionary about it was that Horsburgh removed from the old pilots the many narratives of individual voyage experiences, and provided synthesis based upon general patterns and information. His main concern was to ensure safer passages in the South China Sea for the British Admiralty’s ships, so over the years he methodically collected information that brought him the admiration of Dalrymple, whom he replaced in 1810 as hydrographer of the East India Company (Cook 2002: 135).

46For sailing along the coast of Vietnam, Horsburgh’s instructions in The India Directory were largely based on the work of the Frenchman, Jean-Marie Dayot, a naval officer in the service of Nguyễn Ánh between 1790 and 1795. Like a few other Bretons, Dayot played an instrumental role in the ascension to the throne of the sole Nguyễn descendant. In charge of the Cochinchinese Navy on behalf of the king, Dayot commanded his two French vessels Donnai and the Prince de la Cochinchine. He also contributed greatly to the training of the Nguyễn Navy and above all, distinguished himself by controlling the battles in the bay of Quy Nhơn in 1792 and 1793, during which the Tây Sơn enemy fleet was destroyed (Cosserat 1917: 178-183).

47During his five years of service, as part of his military campaigns, Dayot sailed up and down the coast and took the opportunity to survey and take precise notes and soundings. Supported by the King and excellent navy officers, Dayot worked in perfect conditions to chart the coast as methodically and systematically as possible. His work is exceptional.

Being employed in the King’s navy, and commanding the European vessels that consisted of the main force of his army, I was obliged to take part in each campaign, which usually lasted from the month of May until the month of October. During five years, I covered the coast from south to north and from north to south with the army. […] The time spent at anchor was used to take measurements, for this, the King had affected for me two large boats with oars, and two sailing boats that were only dedicated to this task and that never left the vessel I was aboard. I was assisted in my work by officers instructed and industrious who put in all the zeal possible, so that there is not one point that was not surveyed in relation to another in five or six different positions by the different vessels and boats. When the season forced us back to Saigon, the King gave me the liberty to dispose of the division as I wished and to sojourn in all the ports as long as I estimated necessary to map them, therefore, there is not one port I did not cover from point to point, from rock to rock, while two other boats were occupied in taking soundings in all corners and nooks; as such, I can insure that there are very few ports that were surveyed as scrupulously and with more exactitude than the ones one will find in my map collection. (Dayot 1807: fl. 6r-6v)

48Dayot composed the most accurate instructions for navigating the coast of Vietnam to have ever been written, complemented by eleven exceptional and accurate maps, drawn by his brother Felix. After some disagreement with the King, Dayot left for Macao in 1795 and then settled in Manila (Cosserat 1917: 178-183) where he finalized his considerable hydrographic work. In 1807, he entrusted his maps and instructions to Captain Felix Renouard de Sainte Croix, who deposited them with the French Dépot de la Marine upon return from his campaign in 1808. The maps were subsequently engraved and published in 1818 in the Neptune du grand Océan et de l’Archipel de l’Asie (Palsky 1989: 65). However his instructions and his maps circulated widely amongst shipping circles before being published, Horsburgh having had an early copy in 1804. In 1818, Captain Achille de Kergariou, on a political survey mission to the Nguyễn imperial court brought along with him Dayot’s manuscript, with the intention to verify the value and validity of the instructions. In a letter sent to the Minister of the Navy (dated 20 March 1818), he commented on the descriptions of Dayot and praised his accuracy and savoir-faire. “I could not provide enough praise to the exactitude with which all these lands are presented on his maps” (Maybon 1920: 33).

49Other navigators such as Auguste Borel in 1817, Louis Rey in 1818 and 1819, the American John White in 1819, and the Baron de Bougainville during his circumnavigation of the globe between 1824 and 1826, also noted in their accounts and journals that they used Dayot’s maps and instructions, so did Captain Dangerfield of the John Adams, on which John Crawfurd travelled in 1822, as well as the French Captains Dumont D’Urville and Laplace in 1830.

  • 31 Dayot used very rudimentary techniques. He did not have chronometers and only worked the coast syst (...)

50For the first time, the information available since the Neptune Oriental was appropriately upgraded, the coastline was updated with new toponyms and new sets of coordinates, and the instructions for navigation along the coast was continuous and very detailed. No foreigner knew or understood the coast and the local sailing conditions better than Jean-Marie Dayot. The most important anchorages were accounted for with mentions of the nature of the seabed, with depths that took in account the effect of the tides and with descriptions of the sandbanks that could cause problems at low tide. What makes Dayot’s work even more remarkable is that he managed to chart the coast with extreme precision despite lacking the appropriate instruments.31 The geographical extent of the maps and instructions available was however restricted to the coast from the Mekong up to Pulo Canton only. At the time during which Dayot gathered soundings and observations, the region of Faifo and Hue was still under enemy control of the Tay Son. Thanks to the earlier map of Le Floch de la Carrière and to a “British map” (the previously mentioned map drawn by Barrow and Parrish in 1793), his brother managed nonetheless to outline the coast up to the bay of Turon.

  • 32 When he exposed his project to the King “he felt so well the usefulness of such a survey that he ga (...)

51As Dayot plied the coast and gathered information from fishermen of the provinces of Phú Yên and Bình Định, he soon realised that the extent of the Paracels islands was not as large as commonly thought. While on duty at the court, he discussed with Nguyễn Ánh the value of accurately charting the area, an idea that was wholeheartedly welcomed.32 Unfortunately the project never took place. His maps and expert nautical instructions for the coast of south and central Vietnam however continued to guide navigators sailing along the Vietnamese coast up to the 1880s.

52When James Horsburgh succeeded Alexander Dalrymple as hydrographer of the EIC, he pursued Dayot’s project of charting the Paracels and Spratlys, and collected the results of numerous surveying expeditions in the South China Sea, most of which were then conducted by the Bombay Marine. The passage between the Paracels and the Macclesfield Bank became safer when Lieutenant Daniel Ross of the Bombay Marine succeeded in 1807 in fixing the position and charting with precision the St-Esprit and Helen Shoals. Horsburgh’s first complete charts were printed between 1821 and 1823.

53The navigation methods in the early 19th century were also much improved, allowing better precision and accuracy, determined thanks to chronometers, one of the world’s first ever patent distance logs, systematically marked and calibrated sounding lines, and the trigonometrical system of surveying from base stations (Davies 2015, pers. comm.). By then, ships, knowledge and skills had much improved, and it became possible to make the end run to and from China without stopping. The route now passed seaward of Pulo Condore then to Pulo Sapata, N towards Cape Varella, staying in the offing, then N using islands/landmarks to Pulo Canton or Champello to turn right for a landfall check on S Hainan, then to Macao, or, to sail from Pulo Sapata to the Macclesfield Bank as advised earlier. The crossing of the South China Sea was now less arduous and dangerous than in Dampier’s time.

Mid-19th Century New Navigation Trends

54Changes occurred in terms of navigation in the 19th century. The shift from sailing ships to steam-powered vessels made journeys quicker and cheaper, and their shallow draft allowed a new manoeuvrability in Asian harbours. The monopolies of the European East India Companies shifted, introducing a strenuous competition which imposed needs to be faster. Fast American and British clippers transporting tea or smuggling opium, and large sailing ships carrying bulk cargo between Asia and the west, plied the South China Sea. The 1869 opening of the Suez Canal marked a change in the traditional routes between Europe, East, and Southeast Asia; overall they were shorter. From a technological point of view, marine chronometers, in development since British John Harrison’s first attempts to build one in the 1760s, were in wide spread use in the 1880s to help determine longitude. To determine one’s position was now a matter of science, namely of measurement and precision, and no more relying on vernacular knowledge, collected rutters and itineraries, and on charts that exposed what was not known as much as what was known with certainty.

55From the mid-19th century, the overall understanding of the climatological conditions of the oceans also improved. This knowledge was notably formed on the basis of the remarkable work of Matthew Fontaine Maury of the US Navy who in the 1850s produced very detailed compilations of meteorological data found in ship logs. As Dayot had pointed out in his instructions, understanding the climate, currents and the local winds, and the alternation of land and sea breezes was fundamental for sailing along the Cochinchina coast. He had learned from local pilots that it was possible to sail in the contrary directions of the monsoon by benefiting from local land and sea breezes, as well as from the tidal currents to make way, and this local practice was taken up by audacious western captains, pressed by time and motivated by profit. By the 1860s, to run down the coast against the monsoon was a necessity. Tea clippers now made their passage south during the SW monsoon and travelled from China back to India or Europe as fast as possible (MacGregor 1983: 22).

  • 33 Although the bulk knowledge was nonetheless still copied from Horsburgh, there was a lot of new inf (...)

56In the mid-1860s, Alexander Findlay published the first editions of his Directory for the navigation of the Indian Ocean (Findlay 1866) and of the Directory for the navigation of the Indian Archipelago (Findlay 1869), in which he included the latest data produced by the Admiralty’s Hydrographic Office on the South China Sea. There were great advances and changes in the quantitative ways in which data was gathered and then collated and analysed.33

57In August 1858 the fleet of Rigault de Genouilly left Canton to set siege in the Bay of Turon. The Spanish reinforcements under the command of Coronel Mariano Oscariz joined them from Manila. This offensive marked the beginning of a new era in the history of Vietnam and led to the signing of the Indochinese Protectorate in 1863. From then onwards, the French put much effort into mapping out their new colony of Cochinchina (region of the Mekong). Nautical charts were composed in tandem with the French conquest. Between the 1870s and 1880s, various military campaigns and geographical expeditions, as well as waves of hydrographical engineers, worked on the creation of a good cartographic baseline (Rouget 1906). The 1883 declaration of the Indochina Protectorate, with the annexation of Annam and Tonkin opened a new period of scientific and cartographic projects.

58In 1884, Commander Louis Dartige du Fournet published a new set of updated instructions, Instructions nautiques sur les mers de Chine (Dartige du Fournet 1884), based on several instruction manuals from the Spanish, British and Germans. It was thanks to the latter that the most accurate chart of the Paracels finally appeared in 1885. For the instructions relative to Cochinchina, Dartige du Fournet used the latest works of various French colonial hydrographical engineers and navy officers (Dartige du Fournet 1884: vii). Over the course of a year (between 1892 and 1893), on board the Comète, a war vessel propelled by sail and steam, Louis Dartige du Fournet sailed numerous times along the coast. The passages were always fast and in all directions, regardless of the prevailing monsoon winds. In April 1893, Dartige du Fournet crossed from Cape St-James to Hong Kong in four days, shooting through the Paracels:

The direct route passes through the Paracels, area that used to be unknown, quite feared, a true ship cemetery where one would go only trembling and only if absolutely necessary. Nowadays still, despite the complete works done by German hydrographers, the area has kept its bad reputation and most of the vessels avoid the Paracels keeping respectable distances.
The weather is so beautiful, so clear, that we do want to avoid a useless detour, so we decided to go through, and pass through the islands. We can distinguish them very neatly as soon as we are less than four miles away. They are really low, just at water level, surrounded by coral reefs that rise from the depths, so that the soundings do not indicate anything as we approach. It would be folly to navigate there at night. (Dartige du Fournet 1897: 127)

59This journal entry shows the changes in the experience of navigation in the South China Sea, driven by time and profit, or military orders, despite the apparent danger still waiting on the reefs. It also shows that by the end of the 19th century, with the limits of Paracels and Spratlys fairly clearly established, the main navigation route could now stay clear of the coast of Cochinchina as possible. 19th century western captains and navigators relied on instruments and a very precise and quantified way for finding their position at sea. There was no need to come nearer to the coast to identify markers anymore, because they worked to exact calculations of speed, time, distance, bearing and depth, moving between positions triangulated geographically. Their vessels autonomy did not require them to come close to the coast on a trip between Macau and Hong Kong either, and meteorological knowledge helped to understand and anticipate changes in the workings of the winds and currents. The regular voyages of steamship companies which connected China and Europe contributed to a new sense of time and dynamic in the South China Sea. After more than three hundred years of sail and sweep, the trans-Asiatic ships that cut across the sea initiated a new era in navigation while, French warships controlled the Vietnamese harbours, patrolled the coast, and French railways and roads initiated new transport trends in what became then Indochina.


60This paper has gathered together the numerous endeavours of various European nations to tame the perilous waters of the South China Sea, and demonstrates a continuous build-up of knowledge of its western coastline. The improvement of navigation in the China Seas was primarily due to a succession of recurring voyages and to the diligence of captains and hydrographic crews who took on the habit of keeping log books since the late 16th century. From the early instructions of the Portuguese, the European charting of the region grew in precision. The secretive Dutch furthered the Portuguese material, while the English and the French used these resources to develop in tandem the most detailed sets of instructions. The 18th century Neptune Oriental composed by d’Après de Mannevillette was a major keystone in this great venture. Similarly, Horsburgh’s India Directory became in the 19th century, a milestone in the evolution of nautical instructions and a model from which all subsequent pilots and sailing directions descend. For the navigation along the coast of Vietnam however, it is Jean-Marie Dayot who deserves the most recognition. Having spent several years at the Nguyễn court, sailing with Cochinchinese naval officers, and having discussions with local fishermen and sailors, Dayot made the most of this local knowledge and his local experience to produce the most complete and accurate instructions and charts ever created for that coast.

  • 34 For a discussion on Nguyễn Ánh’s particular interest in naval matters see Mantienne (2003).
  • 35 This, however, is seldom reflected in the nautical instructions. Even if it is widely acknowledged (...)
  • 36 This paper only hints at the potential role of local seafarers. To assess their implication in west (...)

61Dayot could pursue with diligence the production of charts and nautical instructions because of the benevolence and passion for maritime matters of Nguyễn Ánh34 and his dedicated marine officers. This is one example that highlights the underlying role of the Cochinchinese in facilitating the European charting of the area. Southeast Asian seafarers contributed to safer navigation in the South China Sea for the Europeans by piloting their ships, by sharing a few concise maps, and above all, by disclosing travel itineraries and information on where to go (or not to go) and on the hazards to avoid. Over time, European ships followed the same routes used time and time again by local and regional traders and travellers. The reliance on local sources is implied in the works of Dalrymple35 and Dayot, even if the processes of using or of laying down hydrographic data was different. The concept of determining one’s position and the needs of accurate and precise data to navigate (such as exact positions, bearings, distances, depths) were not the same between the westerners’ and the locals’ approach to navigation, but a non-quantitative knowledge of the area was passed on nonetheless.36

62While indirectly hinting at the important role of local seafarers, this paper provided primarily a context to understand the circumstances of the creation of European navigational knowledge, and the conditions by which the Europeans arrived to Cochinchina. As European seafarers expanded into Asian waters, there was an increased progression in acquiring precise navigational data from the 16th and 17th centuries non-quantitative “eye-balling” navigation of the first Portuguese seafarers to the 19th century measured mathematical processes of hydrographic teams of the Bombay Marine. By the 19th century, accurate knowledge was necessary and the qualitative nature of the local knowledge and experience did not meet the new standards necessary for a 19th century captain’s mindset anymore. The captain now needed precision in measurements, depths and accuracy. Until then however, European charts can be seen as the end result of both local non-quantified knowledge and of a western method and approach of quantifying geographical and hydrographical knowledge. In this way, the European charts can be used as a point of entry from which to discuss navigation in the South China Sea and explore the hydrographical knowledge of the Cochinchinese, their navigational tools, and seafaring traditions. Since there is little written on local seafaring, this chronology of the emergence of European navigation in the area is an initial step in comparing how navigation along that coast was conducted.

  • 37 Even so, Horsburgh often noted that the Inner Passage was preferable, particularly in the NE monsoo (...)

63This paper also introduced an important segment of an international maritime route that passes along the Vietnamese coast and showed that this same route was used repeatedly over the centuries, and was often considered as the preferred and safest route, regardless of nationality and time. For more than three hundred years, European ships sailed along this coast, as did the regional Chinese, Japanese, Malay and Cochinchinese seafarers. Regardless of the importance of the harbours and of the types of goods available in Cochinchina, it is clear that to sail past the coast of Vietnam was the safest and most common route to China and back. This highlights the natural inclination of this coast to attract and bring sailing ships of both westerners and locals alike. The fear of the Paracels and of the unknown dangers of the open seas (and for Europeans, the absence of precise navigational data) greatly contributed to the permanent use of this maritime route. Even when instructions of Pimentel or d’Après de Mannevillette suggested a passage away from the coast (to the Macclesfield Bank), it is not until the mid-19th century that crews who did not have specific reasons to go to Cochinchina sailed away from the coast.37

  • 38 Compare for example the mapping of the central coast of Vietnam, with the Gulf of Tonquin or the co (...)

64Interestingly, this paper also demonstrated that nautical instructions for Vietnam were seldom upgraded. Despite the fact that Turon/Faifo was an important international entrepôt visited by European trading companies, missionaries and envoys since the 16th century, it is only in the late 1790s that attention was drawn on accurately charting its harbour. Similarly, maps and descriptions for Pulo Condore, which was a crucial navigational landmark for ships from Europe, China, the Philippines and the Malay Peninsula, did not improve since the first edition of the Neptune Oriental. When looking at regional history and the intense sequel of exchanges between the Nguyễn and the Europeans, the resulting poor mapping of the coast of Vietnam appears as a startling anomaly. Yet, stepping back from this nation-centred perspective and considering the development of navigation of the South China Sea in a global context suggests that the charting of the seas occurred more intensively in other places because the commercial (and political) imperatives and interests were focused elsewhere.38 This limited updating of charts and instructions can also be seen in parallel with the changes in navigation per se over time. Once the main hazards and the most conspicuous landmarks had been identified, it was preferable (and possible) to sail at a certain distance from the coast, and there were fewer needs to survey its waters in more detail and to come closer. The type of coast also had some impact on the feasibility of hydrographical surveys.

  • 39 See the works of Li Tana, Charles Wheeler, Momoki Shiro, and to some extent, about maritime trade a (...)

65Overall, this European perspective on the coast of Vietnam has highlighted the correlation across time between the ecology of the coast and of the South China Sea, with the history of navigation in Vietnamese waters, and leads to see how it would be insightful to develop it further in order to put forward aspects of Vietnamese maritime history. A few scholars have hinted at this maritime connection39 but it deserves to be further investigated, and especially from a Vietnamese perspective. The study of Vietnamese seafaring capacities, boatbuilding, and maritime activities are topics that would contribute to discuss the fact that Vietnam’s “indebtedness to the sea is uncontested” (Kleinen 2011: 292). This paper is a small step in contributing to our understanding of Vietnam’s maritime conjuncture across time and has hopefully set the basis for further approaching the experiences of sailing along its coasts, by foreigners and Cochinchinese alike.

Haut de page

Note de fin

1 The terms “Champa” and “Cochinchina” refer to different geographical regions at different periods of time. “Cochinchina” between the 17th and the 19th centuries refers to central Vietnam, from the province of Khánh Hòa [12°49'N; 109°22'E] up to the capital city of Huế [16°27'N; 107°34'E], where the Nguyễn Lords settled and thrived. After 1867, the term “Cochinchina” refers to the French protectorate encompassing the region of the Mekong, and in turn, central Vietnam in 1884, became the French protectorate of “Annam”. When discussing “Cochinchinese sailors”, it is used as a general term to distinguish the sailors of central Vietnam, under Nguyễn rule during that period of time, from the Cham, and the Đại Việt sailors from the north, under the Trịnh rule. These terms are used to keep consistency with most of the European source material on which this paper is based.

2 “Champa” refers to the extent of the former polity of Panduranga in the 16th and 17th centuries, from the capital of Phan Rang [11°33'N; 108°59'E] to Phan Thiết [10°55'N; 108°6'E] (Dharma 1989: 130; 131). After 1653, the capital moved to Phan Rí [11°11'N; 108°33'E], which marked the then northern limit of Champa until the 19th century (it was annexed by the Viet in 1692, but remained mostly Cham, although under Viet administration.). When the fleet of Fernao Perez de Andrade stopped, Champa still included the polity of Kauthara, whose northern limit until 1611 was the Cù Mông Pass [13°40'N; 109°10'E].

3 Chinese author Liu Xu, in his Old history of the Tang Dynasty, mentions that Chinese officials controlled the barbarians in the south, “and those who dwell on the great sea […] arriving in boats after traveling unknown distances […] bringing goods by the Jiaozhi route as they have done from the time of Emperor Wu of the Han” [140-87 BCE] (Dutton et al. 2012: 16). This is just one instance that hints at the fact that the Jiaozhi (Vietnam) route was frequented since at least the Han Dynasty. See Manguin (1976) and Wade (2013) for an overview of the Chinese itineraries that sail past the Vietnamese coast since the 1st century CE.

4 About cartography in Southeast Asia, see Fell (1991), Schwartzberg (1994) and Suarez (1999). About Vietnam more particularly, see Manguin (1976), Quach-Langlet (1989), Whitmore (1994).

5 In the 16th century, the navigational technology did not allow navigational precision greater than plus or minus 20 to 40 miles (Davies 2015, pers. comm.).

6 See Manguin (1972: 74, 109).

7 As seen in twenty-four rutters for the region, in the series entitled Advertencias para a navegacae da India, compiled between c.1564 and c. 1655.

8 These toponyms can be related to the Malay words barahla, pulau, sisir, kambing, etc. (Manguin 1972: 56-59, 67, 69, 83, 145). Padaran comes from the name of the Cham polity Pangdarang (Panduranga) (Chebaut-Me Mougamadou 2013: 24). Pulo/Poulo/Poolo is the transcription of the Malay word “pulau” meaning island (Cù Lao in Vietnamese). The use of these toponyms in European languages may reflect the fact that most of the pilots employed by the Portuguese, Chinese, and others, were Malay-speaking people.

9 The Scorpion’s Tail was also an imaginary bank, believed to be connected to the Paracels; “It is from there that begins the dangerous channel which has to be passed to go to China or come back. From this “Ravo” until sixty miles beyond Poukatan [sic], there is a continuous series of shoals, on which many vessels get lost every year” (Gemelli Careri in Cadière 1930: 307).

10 See Manguin (1976) who briefly considers the question, as well as some famous Chinese rutters that describe the Paracels as a great danger, and Manguin (1972) for the experience of Portuguese sailors.

11 Note that if the Paracels were not surveyed and charted until the 19th century, it is also because low-lying steep to coasts are more difficult than other areas to chart. Ships would not risk coming too close to gather data. Similarly nearing the Mekong area, which is on a shallow to coast, with a very low hinterland, it was difficult to determine one’s position before being in shoaling waters, full of sandbanks; no ship would try to gather precise data there until the end of the 18th century.

12 In the 15th century, King Joao II of Portugal tried to enforce a “policy of secrecy” (sigilo) to keep other nations from exploiting Portuguese new maritime and navigational discoveries. Despite a secrecy attempted, there were “leaks”, theft, and passing on of knowledge, the most famous example being the theft of the Cantino Planisphere in 1502. Linschoten “copying” the material of the archbishop of Goa is another example.

13 See the rutters of the mid-17th century in the Advertencias (Manguin 1972: 91-103; 107-120; 122-125). Note however, that this does not mean that navigation was precise or accurate. Indications every 20 to 30 miles do not provide great accuracy, and systematic knowledge of tides was rudimentary until the 19th century.

14 The rutters in chapters 20, 24 and 44 of the Reys-gheschrift are copies of some Portuguese rutters that can be found in the Advertencias (Manguin 1972: 62).

15 Lucas Janszoon Waghenaer’s Spieghel der Zeevaerdt (1584/1585) and Het Licht der Zee-vaert of Willem Blaeu senior (1608) constituted the premises of this new tradition of sea pilots which took a new step beyond the earlier Portuguese data sets and portulan tradition. Waghenaer’s Spieghel der Zeevaerdt was the first true combination of navigation instructions with nautical charts. It was counterfeited in English in 1588 under the name “Mariner’s Mirror” but since then the term “waggoner” has been used in 17th century English to refer to sea-pilots.

16 For example, Linschoten obtained information about the navigation between China and Japan from his friend Dirck Gerritz. Waghenaer included in the 1602 version of his Thresoor der Zeevaerdt instructions about the Sunda Straits region, based on the expedition of Jacob van Neck and Wybrand van Warweick (1598-1599).

17 Dutch navigation instructions were not in wide circulation, in contrary to the Portuguese instructions in the previous century or the English instructions of the following century. It was not only secrecy but rather exclusivity which provided the VOC charts with a secret coat, as Zanvdliet demonstrated, “instructions to submit charts and instruments after the return of a ship had less to do with secrecy than with improving efficiency and guaranteeing the efficient re-use of expensive supplies […] Manuscript charts and maps almost always circulated in smaller numbers than printed maps and charts; their numbers were lower and their prices higher. As exclusive items, they were collected and prized” (Zandvliet 1994: 1462).

18 Asian ships followed the same route, however tended to veer to Hainan from Pulo Canton, as attested by numerous Chinese itineraries and the famous Selden Map (see Wade 2013).

19 An estimated 247 VOC ships were lost in the 17th and 18th century in the East Indies (Hol et al. 2012), including the Grootebroek (1634) and the Arion (1714) in the Paracels. No extensive survey has been conducted so far as to estimate the number of shipwrecks that lie along the coast of Vietnam.

20 In the 1660s, four junks per year sailed between Manila and Cochinchina (Li 1998: 76). Several instances in the Spanish Archivo General de Indias, Audencia de Filipinas, relative to the death of Governor Gomez Luis Perez das Marinas in 1593 and the subsequent missions that went to Cochinchina and Cambodia, hint at a crossing between Manila and the south of Vietnam. De Morga, recounting the history of the Spanish in the Philippines mentioned Juan de Mendoza and his pilot, Juan Martinez de Chave, going to Siam in 1593, “over the shoals” (Morga 1890: 135-136). Later, discussing their return to Manila, he noted they sailed up to Cochinchina because the ship was too battered “for crossing the shoals to Manila” (ibid.: 184).

21 It was still based on the idea of simultaneous observation of a celestial event from two different locations. The value they obtained for Lop Buri was more than 3° too far east, therefore about 180 nautical miles off. This shows the levels of accuracy of 17th century navigation.

22 D’Après de Mannevillette was navy officer until 1759 and then appointed chief of the Dépot des cartes et plans de la Compagnie des Indes in Lorient in 1762.

23 He produced another map, dated to 1755-1759, but published in 1787. It represents the same portion of the coast (ca. 1: 360 000) and comprises a profile of the citadel of Huế, depicts native ships and provides some instructions.

24 M. Didier was an engineer of the King, and commander on the island in 1721 to 1722, when the French had intentions of settling there. His chart was only slightly upgraded in the second edition. In fact, it is William Dampier who drew the very first map of this island (Dampier 1697: 389-390), but it was really inaccurate.

25 Huddart’s version, while repeating the content of the Neptune Oriental, re-organised the layout, the break-down of the routes, the order of the descriptions, and added new sections and many detailed narratives of other Captains' experiences as well as the meteorological data from Dunn. Whilst Herbert’s and Dunn’s are copies of the content of the Neptune Oriental, Huddart’s Oriental Navigator is truly an adaptation and more systematic than what existed before. It was followed by Wright’s improved version A New Nautical Directory for the East-India and China Navigation (1804).

26 They all repeated the same mistakes. For example, the inexistent island “Pulo Cambir de Mer”, which appeared in 1665 on a Vingboons map, was still present in the Oriental Navigator. Until the 19th century and systematic surveying, there were no ways to eradicate errors. Any report of a danger was as good as any other so all dangers were retained into the record, good or bad. When possible, errors and wrongly positioned elements were eliminated, but if not enough information as available, these were still included (Davies 2015, pers. comm.).

27 The Macclesfield Bank is a easily recognisable, very large (6 448 km2) and submerged bank (shallower depth is c. 9 m, in the SE corner, where HMS Macclesfield went briefly aground in 1804). As such it was a godsend for the mariners.

28 In subsequent navigation manuals such as James Horsburgh’s India Directory, this “east of the Paracels passage” is usually termed the “outer passage” (in opposition to the “inner passage” or traditional route which borders the coast) (Horsburgh 1843: 292-293), while the “eastern passage” refers to the route that follows the western coast of the Philippines from the Sunda and Macassar Straits (ibid.: 542). Now that navigation tools are more performing and accurate, it is the most common passage to link Singapore to Hong Kong.

29 Pirates in Southeast Asian waters have always been a problem; until the late 17th and early 18th century, in southern and central Vietnam, the Cham were renowned for tricking and luring ships into their bays. By the mid-18th and mid-19th centuries, Chinese and Vietnamese pirates were a good reason for keeping off the central and northern coasts of the country. D’Après de Mannevillette, advising not to come too near Cochinchina/Champa, recounted in the Neptune Oriental the full story of the Galathée, a ship, tricked and trapped in the Cham harbour of Phan Rang for over a month in 1720, in order to demonstrate how perilous it could be to linger along that coast (1775: 435-449). The Selden Map also provides information about pirates or locations where the tradition of having ownership on “salvaged” wrecks is applied.

30 The English Pilot has about 30 landmarks whilst the Neptune Oriental has about 40.

31 Dayot used very rudimentary techniques. He did not have chronometers and only worked the coast systematically, taking precise notes and observations. As he noted in his introduction, he surveyed the coast over five years and paid attention to each location from five to six different positions. When possible, he also cross-referenced his measurements with the ones of Mr. Olivier, engineer, commanding one legion of the King’s Guard, on land.

32 When he exposed his project to the King “he felt so well the usefulness of such a survey that he gave me the order to prepare for it. He gave the most precise orders so that I would be supplied all I would judge necessary from his arsenal to facilitate this operation. I had ‘carte blanche’ […]”. Twice, the project was postponed because of the Tây Sơn threat and the needs of Dayot to attend the arsenal and the royal fleet. After a misunderstanding with the king, Dayot left the court. In 1809, Dayot and his family drowned in a cyclone in the Tonkin Gulf.

33 Although the bulk knowledge was nonetheless still copied from Horsburgh, there was a lot of new information, such as the new lighthouses, and differences in the way the data was organised. By the later editions of the India Directory, a new “standard” for pilot layout had been elaborated, which was subsequently followed by the British Admiralty and then all other sailing directions. The Admiralty’s China Sea Directory began life following the 8th edition of Horsburgh in 1864 and was first published in 1867. Once Horsburgh ceased publication, the non-Admiralty route was taken up by Alfred Dundas Taylor, expressly compiled “for the Guidance of Commanders of Steamers and Sailing Vessels” (Dundas Taylor 1874). Findlay’s work went through three editions, the last in 1889. In 1895 the Admiralty’s Ocean Passages for the World was published for the first time. With routes for both sailing and steam powered vessels, it became the bible for offshore route planning. The first volume of the US Hydrographic Offices Asiatic Pilot came out in 1909, and the last of the six volumes in 1919.

34 For a discussion on Nguyễn Ánh’s particular interest in naval matters see Mantienne (2003).

35 This, however, is seldom reflected in the nautical instructions. Even if it is widely acknowledged that local pilots where hired, only a few like Dalrymple clearly informed the reader when his observations were drawn from local pilots. For the Paracels plan he wrote “The Paracels are laid down from a plan made by a CochinChinese pilot, […] I found the southern islands of the Paracels agree tolerably with the observations of some English ships, by whose accounts I determined their position” (Dalrymple 1786: 9). On the cartouche of his map of Borneo, and in his memoir, he also clearly noted that it was based on “a sketch of Dato Saraphodin and from a chart of Noquedah Koplo, a Javan Chinese who came up that coast in 1761” (Dalrymple 1786: 3).

36 This paper only hints at the potential role of local seafarers. To assess their implication in western navigation it would be useful to first assess their navigation tools and concepts. For this, see Davies (2012). Furthermore, navigation for a local fisherman on a small (<10 m) boat in familiar waters and for the European captain of a cumbersome East Indiaman in unknown waters was very different and cannot be compared.

37 Even so, Horsburgh often noted that the Inner Passage was preferable, particularly in the NE monsoon, and during most of the SW monsoon (Horsburgh 1843: 292; 298). Between 1700 and 1799, about 800 EIC ships went to China (Farrington 1999). A detailed analysis of the content of their logs would allow determining with precision whether they sailed along the “Eastern passage” or whether they preferred to sail along the coast of Cochinchina, and at what times.

38 Compare for example the mapping of the central coast of Vietnam, with the Gulf of Tonquin or the coast of Cambodia which were, during the 17th-19th centuries barely outlined because of the lesser economic and political interests for the region. Portuguese sailors tended to sail straight from Melaka to Pulo Condore and then up the coast, following the traditional route, and hence not stopping in the western portion of the region. Likewise, the lesser presence of Champa on the Dutch maps most likely reflects the absence of commercial ties between the Cham polities and Batavia. This started to change in the 1790s.

39 See the works of Li Tana, Charles Wheeler, Momoki Shiro, and to some extent, about maritime trade and activities in Nguyễn Cochinchina, Frederic Mantienne, and Nguyen The Anh.

Haut de page

Pour citer cet article

Référence papier

Charlotte Minh Hà Pham, « European Navigation, Nautical Instructions and Charts of the Cochinchinese Coast (16th-19th Centuries) », Moussons, 27 | 2016, 101-129.

Référence électronique

Charlotte Minh Hà Pham, « European Navigation, Nautical Instructions and Charts of the Cochinchinese Coast (16th-19th Centuries) », Moussons [En ligne], 27 | 2016, mis en ligne le 17 mai 2016, consulté le 27 mars 2017. URL :

Haut de page


Charlotte Minh Hà Pham

Charlotte Minh Hà Pham is a PhD Candidate at the Asia Research Centre (Murdoch University), in Australia. She is funded by an ARC Linkage Grant, as part of a Major Collaborative Research Initiative (MCRI) project anchored at the Indian Ocean World Centre (IOWC), McGill University, Canada. Her initial fieldwork was conducted thanks to an École française d’Extrême Orient (EFEO) scholarship in 2010-2011 on the boatbuilding traditions of Vietnam. Her PhD focuses on the maritime cultural landscape of Cochinchina between the 16th and 19th centuries, and includes further research on Cochinchinese boatbuilding, maritime routes, navigation conditions, and seafaring and seamanship.

Haut de page

Droits d’auteur

Licence Creative Commons
Les contenus de la revue Moussons sont mis à disposition selon les termes de la Licence Creative Commons Attribution - Pas d’Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page