Abstract
This article questions the commonly held assumption that the ships of the Dutch East India Company VOC were slower than those of other East India companies. Recently, Solar and De Zwart showed that Dutch ships were slower on outward voyages to a number of Asian destinations during the periods 1770–1775 and 1783–1792. They cited as plausible explanations differences in ship design resulting from constraints imposed by the Dutch shallow inland waterways and the slow adaptation of copper sheathing in the late eighteenth century. Research by the author of this article leads to a critical assessment of these explanations. Moreover, additional new research into homebound voyages from China undertaken by ships of four East India companies, for the periods 1730–1740, 1750–1755, 1770–1775 and 1783–1792, leads to the outcome that – concerning speed – Dutch ships could compete very well with those of the English, Swedish and Danish companies.
The ships of the Vereenigde Oostindische Compagnie (VOC, the Dutch East India Company) have often been characterized as sturdy, but also as heavy, slow and lumbering cargo ships. Ships of the English, French, Danish and Swedish East India companies are usually considered to have been faster sailing ships.
1
In their classic work on the ships and shipping of the VOC, Bruijn et al. compared sailing times of different companies. They concluded that, on average, the ships of the VOC took longer to cover the same route than the ships of these other companies.
2
Contemporaries made critical remarks regarding the speed of the Dutch ships in the eighteenth century. In a meeting in August 1775 of the Heeren XVII (the 17 directors of the VOC), Willem V, stadholder of the Dutch Republic and honorary chief director of the VOC, explained why the English and French ships crossed the equator faster than the ships of the VOC: the English and French operate with more and lighter sails, are lighter rigged, and cross (the equator) faster; the Dutch (ships), built for heavier loads and with a heavier rigging, sail quite well with a fresh breeze from astern but poor in light winds, and so it usually takes them a long time (. . .).
3
In their response, the Heeren XVII promised to: strip the ships from unnecessary rigging blocks, and to provide the ships in addition to the usual sails also with a sprit-topsail and a top-gallant-sail and to attach wider sheets to the staysails . . . .
4
This discussion shows that the VOC was aware of the fact that its ships were (at least under certain conditions) slower than the ships of the English and French companies. Apparently, this lack of speed did not matter too much when it concerned the shipping from the East Indies to the Dutch Republic of non-perishable goods for which the VOC had an effective (near) monopoly, such as cloves, nutmeg, cinnamon and pepper. It may have mattered though on the outward voyage, because a prolonged voyage increased the risk of scurvy or contagious diseases, such as typhus, which were rife during a number of years, most notably in the early 1770s. 5
Matters were different in the competitive tea trade from Canton in China, a trade that during the eighteenth century became increasingly important to the India and Asiatic companies based in northwest Europe. Tea was a much sought-after commodity in that region during the second half of the eighteenth century, and getting tea to the European markets earlier than competitors was an important factor in procuring a good price. A lengthy transport of tightly packed tea caused an unavoidable deterioration in quality, while the repacking and storage of the tea in Europe led to a further loss of quality. It was therefore of great importance to arrive in Europe in time for the tea auctions, which always took place in the autumn, preferably ahead of rival importers. By doing so, it was not only possible to fetch a good price for the tea, but it also helped to avoid a long period of storage. 6
Peter Solar and Pim de Zwart have recently considered a number of reasons for the (assumed) slowness of the ships of the VOC. 7 By comparing the time taken by ships of different East India companies on a number of outward journeys towards Asia, they concluded that the ships of the VOC ‘were generally much slower in getting to Asia than those of other companies’, in particular compared to English ships. In addition, they quoted Bruijn et al., on the fact that Dutch ships were one to two months later in getting back from Canton to Europe, compared to the vessels of the French, Danish and Swedish companies. Part of the explanation given by Bruijn et al. was the relative inexperience of Dutch skippers (captains) on that route. 8 However, that inexperience can only be a valid argument concerning the route between Canton and the Cape of Good Hope (and even more specifically on the route between Canton and Sunda Strait), as Dutch skippers had ample experience of the route between Java, the Cape of Good Hope and the Dutch Republic. For that route there was an extensive set of sailing instructions provided by the VOC, although these instructions came also to be seen – not only in hindsight, but also by contemporaries – as restrictions that prevented skippers from exploring faster routes.
Solar and De Zwart argued that ‘slow ships increased the cost of trading with Asia’ and they pointed to the fact that the VOC ‘still competed with the English and French companies in third markets in Europe’. 9 However, whereas their essay only considered outward-bound voyages, this article extends the research by concentrating on homebound voyages. To be more precise, it looks at the duration of homebound voyages by ships of different northwest European East India Companies between Canton in China and Europe for which speed was quite important. Comparing these voyages is made easier because these companies sailed by and large along the same routes. In addition, by concentrating on the competitive tea trade with China, issues of the VOC’s monopoly on the trade in certain goods and on certain routes are of no consequence. If the Dutch had made incorrect decisions regarding routes or ship design it might have cost them dearly in the tea trade with China.
However, before we consider the duration of the China-Europe voyages, we will look into the arguments put forward by Solar and De Zwart regarding the ‘growing divergence of ship speeds from the 1770s’ and their ‘explanation of the initial gap’. 10
Did ship speeds diverge from the 1770s?
The explanation for a growing divergence was presented as a straightforward one, namely the rapid adoption by the English East India Company (EIC) of copper sheathing for their ships’ hulls. No precise timeframe is given by Solar and De Zwart for the introduction of this innovation, but the wording suggests that it took place in the early 1780s. However, Jean Sutton cited William Hickey (a lawyer employed by the EIC), who wrote in 1779 that ‘scarcely any of our ships were copper-bottomed’. She went on to say that the ‘old owners’ tried as late as 1786 to push a motion through the courts that would prohibit copper-bottomed ships being employed in future. 11 Although ‘new shippers’ and the English supercargas (chief merchants) in Canton were more favourable towards coppering, by 1790 only 22 of the company’s ships had been coppered – out of the 87 ships employed in that year by the EIC. This does not suggest a very rapid adoption of the system of coppering, considering that the British Navy had most of its ships coppered by the early 1780s. 12 However, corrosion caused the Navy a number of problems and even led to the loss of some ships; problems that were by the mid to late 1780s solved by the introduction of copper (or copper-alloy) bolts instead of iron bolts. 13
In the Dutch Republic, meanwhile, the Amsterdam admiralty had independently started experimenting with copper sheathing in 1777, followed by the Rotterdam admiralty in the early 1780s, Enkhuizen & Hoorn in 1783 and Zeeland in 1786. 14 Problems with corrosion delayed a full-scale introduction and by 1795 several vessels in the Dutch navy had not yet been coppered. 15 Considering this, the reluctance of the VOC regarding the introduction of coppering can be seen as precautionary rather than as a sign of conservatism. In a memorandum of 1788, the VOC staff in Canton asked (among other things) for coppered three-deckers in the China trade, and for skippers to be given up-to-date foreign charts. 16 The ensuing decision of the VOC in May 1790 that all yards not using ‘camels’ (which meant that only the shipyard of the Amsterdam chamber was excused, but that the other five chambers had to comply) had to build coppered ships for the China trade, the subsequent 1792 request to have all new ships coppered and the obligation to do so in 1794 show (for a large company with its own shipyards) a rather fast turn-around. It is interesting to note that both in the case of the EIC and the VOC, the staff in Canton played an important part in the acceptance of coppering.
According to Erik Gøbel, the Danish Asiatic Company started coppering its ships from 1792 onwards, at the same time as the Danish navy started coppering its fleet. By 1796, six out of the eight ships employed by the company had been coppered. Although one of the advantages of coppering was assumed to be a greater speed of the ships (contributing to shorter voyages), study of the duration of the voyages undertaken after 1792 does not point to faster sailing times. Apparently, the Danish company did not pursue faster voyages than usual. 17
The speed of the ships of the EIC did increase outward bound between 1783–1792 compared with 1770–1775, but (looking at Solar and De Zwart’s Table 1) only on the outward voyage to Bombay, Madras and Bengal. On the Batavia route, the small number of English ships took slightly longer than before. Direct comparisons with the VOC can only be made for the Bengal and Batavia route and there we see that – although the Dutch were still slower than the English – Dutch ships (despite not having been coppered) tended to become faster as well and that they were actually closing the gap with the English ships on the Batavia route. It may have been the case (as Solar and De Zwart mentioned) that the faster speed of coppered ships made it possible for the ships of the EIC to sail from the South American coast towards the roaring forties from the 1780s onwards, but as we have seen before, the majority of the EIC ships were still not coppered at that time. Moreover, the roaring forties route could also be used (albeit over a shorter distance) after stopping at St Helena (the intermediate stop for the EIC) or at the Cape of Good Hope (which was the obligatory intermediate stop for the VOC). In fact, the roaring forties route had been prescribed as the outward route by the VOC since 1617.
Comparison of sailing times from China as performed by the adapted Delft/Rotterdam design for the second charter (built from 1764 onwards) and by the ships of the second charter built according to the standard design (Delft/Rotterdam ships built before 1764 and the ships built at all other yards from 1744 onwards).
What about an initial gap?
Solar and De Zwart argued that ‘there was no significant trend over the late seventeenth and eighteenth centuries in the duration of Dutch voyages to Batavia’. They referred to Bruijn’s statistics for 1720–1749 and 1750–1779. 18 The fact that ships sailed faster to and from Batavia and Ceylon during the second timeframe of Bruijn’s statistics may well be attributed to the new designs, drawn by the British master shipwright Charles Bentam and effectively introduced in the late 1740s. The extensive statistics in Bruijn et al. regarding the sailing times from the Republic to the Cape show this even more clearly. 19 Before the mid-eighteenth century, the VOC had its ships built according to a design that had been agreed upon in 1697. Adaptations to this design in 1714 (to cope with shallow estuaries) resulted in ships becoming wider and flatter bottomed; this made the ships less seaworthy. This adapted, rather cumbersome 1714-design may have been the cause of an initial gap. Several shipping disasters then made the federal government of the Dutch Republic step in, threatening not to prolong the monopoly of the VOC. This led to the complete overhaul of the standard designs of the VOC by Bentam, which removed the original cause of an initial gap. Of the three classes (‘charters’) of ships designed by Bentam, only the first (largest) and second (intermediate) were built in great numbers. The third (smallest) charter was hardly built at all. 20
The remark that shipwrights ‘took significant liberties’ in their shipbuilding may be applicable to the period before 1742, but thereafter the VOC used technical, measured drawings, which meant that there was (with an existing system of control in place between the different yards of the VOC) little possibility of bending the rules. Contemporary evidence suggests that these drawings were indeed used to guide and control the building process. 21 The new designs had a hold (draught) that was larger than that of the preceding designs. This must have been why the Delft and Rotterdam chambers tried to adapt the design in 1763, although only concerning the second charter. 22 Delft and Rotterdam were granted permission in 1764 to lessen the draught of these ships by 1½ feet and to compensate for that by building the ships wider and blunter. This happened against the wishes of the master shipwrights of the other chambers, one of the arguments being that it would mean a return to the unsatisfactory 1714 designs. Considering the fierce opposition of the other shipwrights, it is highly unlikely that the other chambers followed the Delft/Rotterdam adaptation. 23 Because of the small number of ships built by Delft and Rotterdam, their adaptation can hardly have had much effect on the average sailing time. Moreover, it certainly did not contribute to an initial gap because this adaptation became effective from 1764 onwards, spanning only the last 30 years of the VOC’s existence. 24 The design of the first (largest) charter was not changed, and nor did the other chambers alter the design of the second charter. This meant that these other yards used the ‘slimmer and sharper’ design by Bentam (see the dotted lines in Figure 1) until the demise of the VOC in the 1790s.
Three frame drawings, comparing the frames from the Bentam design for the second charter (dotted line), the Delft/Rotterdam design (the innermost of the two solid lines, in the original drawing in red), and the 1714 design (the outermost of the two solid lines).
Considering the adaptation made by Delft and Rotterdam, it is indeed rather curious (as remarked upon by Solar and De Zwart) that ships from Delft turned out to be faster sailers on the outward route to Batavia than ships of the other chambers. 25 However, from Table 1 it can be seen that the Delft/Rotterdam ships were on average the slowest ships between China and the Cape. It took the ‘flat bottomed’ and ‘more blunt’ Delft/Rotterdam ships about a week longer to sail from China to the Cape, while the voyage from the Cape to the Republic took these ships two weeks longer. However, when we discard the very long voyage of the Rozenburg (which was incidentally the very last ship to reach the Republic from China) then it turns out that – compared to the standard design – these ‘blunt’ ships still needed about a week longer to reach the Cape from China, although they needed almost a week less to sail from the Cape to the Republic. This shows that the adaptation had not much of an effect on the duration of the voyages, at least on the homebound voyages from China. The staff in Canton noted in 1788 that one of the fastest English ships used in the Canton tea trade was the Vrouwe Catharina Wilhelmina, a Rotterdam-built ship of the second charter that had been taken by the English in January 1781, and subsequently coppered to become one of the EIC’s fastest ships. 26
This suggests that, even in its more blunt form, the 40 years’ old 1742 VOC ship-design could still compete with English ships. Whether the fact that ships tended to sail faster to Batavia after the demise of the company can be construed as indirect evidence for the inadequacies of the mid- eighteenth century VOC ship design, seems somewhat questionable. 27 It is as least as likely that the extra speed was made possible by the fact that the increasingly detailed sailing instructions no longer had to be adhered to after the VOC had ceased to exist.
The remark by Solar and De Zwart that the VOC proposed that new charts, based on French work, should be used is somewhat optimistic. 28 It was eventually decided to supply only ships bound for China with those charts. That decision was taken in the wake of the wrecking of a Dutch East Indiamen on an uncharted island in the South China Sea upon its return from China. 29 Considering the rhetorical question whether the prescribed routes would have been maintained if they were suboptimal, one only needs to look at contemporary remarks by skippers and staff of the VOC (such as the staff in Canton and examiners of VOC’s officers), and by the Batavia-based Bataviaasch Genootschap van Kunsten en Wetenschappen (the Batavia Society for Arts and Sciences). 30 Their sometime scathing comments pointed to severe shortcomings in the existing instructions and to the superiority of foreign charts and instructions. The reason for the VOC to hold on to its own charts and instructions may well have been that the company wished to maintain (or even increase) a tight grip over its nautical operations and in particular over its skippers. 31
Homebound from China
We now turn our attention to a number of the homebound voyages that were carried out by the ships of the Danish Dansk Asiatisk Compagni (Dansk), the English East India Company (EIC), the Swedish Svensk Ostindiska Compagnie (Svensk) and the Dutch Vereenigde Oostindische Compagnie (VOC). To be able to make comparisons with the numbers presented by Solar and De Zwart, their timeframes 1770–1775 and 1783–1792 have been chosen, with the addition of the periods 1730–1740 and 1750–1755. 32
Table 2 shows that on the first part of the homebound voyage (the route between Canton and the Cape of Good Hope/St Helena) the Danish and Swedish ships always recorded the fastest average sailing times, even though (just like the Dutch) they had a larger distance to cover than the English ships. 33 Possible reasons for this state of affairs will be discussed below. During the 1730s, on that same part of the route, the VOC ships managed a faster average sailing time than the EIC, which seems to contravene the idea of an ‘initial gap’. During the early 1750s and 1770s, the VOC ships sailed slightly slower than the EIC ships, but the difference is not huge and – relatively seen – it did not increase over time. The picture for the final period is rather remarkable. When we discard the Dutch ships that, having left Canton late in the 1788 season, took more than 50 per cent longer on the route towards the Cape, we see that the average sailing time of the Dutch ships is very close to the time the English needed, even when we also discard English ships that took 50 per cent longer than usual. 34 Moreover, sailing into Table Bay (or False Bay) at the Cape of Good Hope could easily take a few days, which is included in the Dutch sailing times but not in those of the Danish and English; the reason for this is that these ships did not usually enter the local bays at the Cape. If the coppering of EIC ships in the 1780s had an effect, it did not show on the first leg of the voyage. On the contrary, the sailing time for the English was as least as long as the times they managed in the early 1750s and 1770s.
Homebound sailings from China.
Sources: Gøbel, Asiatisk; Farrington, Catalogue; C. Koninckx, The First and Second Charters of the Swedish East India Company (1731–1766) (Kortrijk, 1977); Bruijn et al., Dutch-Asiatic.
The second leg of the homebound voyage (between the Cape/St Helena and Europe) paints a more complicated picture. Whereas the Danish ships made the fastest crossings during the 1730s (and the English the slowest), in the early 1750s the VOC ships recorded the fastest time with the EIC still taking the longest. In the early 1770s, Danish ships were again the fastest sailers, and the VOC and EIC ships recorded almost the same time – when we realize the extra distance the VOC ships had to master to get to the Republic, we may even assume that the Dutch were slightly faster. It is interesting to note that between 1783 and 1792, the Dutch were closing the gap (as far as the total sailing time is concerned) with the Danes. It is only in that period on the homebound stretch from St Helena to Europe that the English ships sailed faster than the Danish and the Dutch ships. Whether that can be attributed to the coppering of part of the EIC ships seems doubtful, considering that no positive effect on ships’ speed was visible on the Canton to Cape/St Helena route. For the same reason it is also very unlikely that the increasing number of larger ships employed by the EIC had an effect, as this would have been visible on both routes. It is also unlikely that the officers of the EIC adopted navigational aids and devices, such as lunar tables and chronometers, earlier than the skippers of the VOC, who only got access to lunar tables from 1789 onwards. 35 That advantage would in all likelihood also have had an effect on the Canton-Africa route. The only possible explanation seems to be that by the 1780s the officers of the EIC had gained more expertise and better knowledge of wind, weather and current patterns of the (Southern) Atlantic and that they – not encumbered by overly detailed sailing instructions – could use that knowledge to good effect.
That brings us back to the very detailed sailing instructions with which the skippers of the VOC had to comply. The instructions for the Canton-Sunda Strait part of the route were less detailed than the remainder of the route. That can be explained by the smaller number of ships that sailed that route and, ironically, by the fact that the charts for that part of the route left something be desired. On the one hand, that made it difficult for the staff in Amsterdam to check whether ships had followed the prescribed route, thus giving the skippers some leeway in deciding routes. On the other hand, it necessitated (as mentioned above) the introduction of French charts for the China-bound ships.
Skippers’ experience and time of departure
Bruijn et al. observed that Dutch skippers and mates had less experience on the China route than their Danish and Swedish counterparts and suggested that this might have played a role in the relative slowness with which the ships of the VOC returned from China. 36 If that was the case, then it probably only played a role on the route from Canton to the Cape. Almost all skippers and mates of the VOC, especially when they had risen through the ranks, had sailed quite a number of times from the Cape to the Republic; moreover, the route between the Cape and the Republic was strictly regulated and ships arriving at the Cape either from China, Ceylon, Bengal or Batavia often had to form convoys on the way to the Republic whereby the slowest of the ships dictated the speed of the whole fleet. This may explain the lengthy duration of the whole voyage to the Republic. 37 When we consider the VOC skippers who sailed at least three times to and from China (see Table 3), there is some indication to suggest that experience paid off, whereby the crux seems to have been the third voyage. After a third voyage, there seemed to be little room for further improvement. Irrespective of the possible slowness of ships or the (in)experience of skippers, the time of departure from Canton was an important factor in the duration of the voyage home. All sailings included in Table 3 left in the favourable months of November until February. That October and March were unsuitable months for a departure from Canton is clear in Table 4.
The effect of experience.
Month of departure from Canton.
Note: This table shows that the ships of the VOC (with only very few exceptions) managed to choose the right months to depart Canton.
It seems remarkable that the Danish and Swedish ships were right from the start faster than their competitors, considering that they were latecomers on the scene without any experience on the route to China; the very first Danish sailing to China took place in 1733/34 and the Swedish company sent out its first ship in 1732. However, both companies engaged experienced officers from the Oostendse Compagnie (Austrian Netherlands), which was active in the China trade between 1715 and 1730/32. The captain on the first (1733/34) and the third (1736/37) sailing of the Danish company was Guillielmo (de) Brouwer, who had been captain of the Marquis de Prié, a vessel set forth by the Oostendse Compagnie in 1727/28. 38 The first mate on the first Danish sailing, and captain on four later sailings, was Zacharias Allewelt, who came from the Dutch Republic. 39 Philip Jacobus Derdeyn, also a former officer of the Oostendse Compagnie, was first mate on the 1737/38 sailing and captain on two later sailings. 40 Likewise, the Swedish company engaged the experienced John Maule (from England) and Pieter Dens (from the Austrian Netherlands) as captains during the first years of the company, as second captains De Clerq, Carpentier and Pierre Brunet, and as second mate Pierre Bedet (all from the Austrian Netherlands). 41 It may even have been the case that the VOC also engaged a former officer of the Oostendse Compagnie as skipper on its first direct sailing to China with the Coxhoorn, because Jan Schull became skipper without ever having been employed by the VOC before. The Coxhoorn (built to the wide and flat bottomed design of 1714) realized a really fast time on the homebound voyage, which strongly suggests that at least some officers had gained expertise regarding the route. 42 Schull sailed one more time for the VOC to China, this time as supercarga. He is not mentioned as captain with the Oostendse Compagnie, so he may have been first or second mate with that company. The supercarga on the first direct sailing had certainly been employed as carga by the Oostendse Compagnie. 43 All three companies therefore hired foreign expertise for commanding their ships and/or for negotiating trade deals in China, and, looking at the sailing times that were realized, to quite some avail.
According to Gøbel, ‘the great majority of common seamen averaged more than three expeditions to China, while petty officers sailed more than four times (. . .) Danish and Swedish captains continued to sail in this high rank decade after decade.’ This means that the Danish and Swedish companies allowed their seamen, particularly their officers, to expand their expertise on the China route; ‘the fifty Danes (that attained the rank of captain) between 1752 and 1807 (. . .) sailed an average of 8.6 expeditions as navigating officers.’ He added that English captains sailed on average only 4.4 times as mate of captain. 44 As we have seen above, Dutch skippers and mates had even less experience on the China route.
Danish and Swedish companies may have been faster in getting home once they had left Canton, but we should realize that on a number of occasions Danish and Swedish ships had to spend the winter in Asia, having arrived too late to reach Canton or to leave Canton in time. Regarding the Danish company, that happened to one ship in the period 1750–1755 (out of 13 ships) and to five ships in the period 1783–1792 (out of 20), usually at the Danish controlled port of Tranquebar in India. Also, on two occasions in the period 1750–1755, Swedish ships (out of the 10 ships that sailed in that period) had to stay during the winter in Canton. This must have meant an increase in costs for the companies involved. After 1755, outward-bound the ships of the VOC with destination China always made a stop in Batavia where a number of crew and all soldiers disembarked and cargo for China was taken on board. When a China-bound ship arrived too late in Batavia, the officials there (who were informed by letters carried by all ships arriving in Batavia during that season about the logistics of the China trade of the same season) provided a replacement vessel. This meant that the complete number of ships destined for China ships arrived in time in Canton and could therefore leave in time for the Republic.
Characteristics of ships
It is unlikely that coppering and the adaptation of ships to shallow estuaries underpinned the growing or initial difference in speed between the ships of the VOC and those of the EIC. Instead, the main characteristics of the ships of the four companies may shed light on why the speed between these ships differed. These characteristics are known for the ships of the VOC because of the standardization that the company favoured and which became more marked over time. 45 They are missing for the ships of the EIC in the periods 1730–1740 and 1750–1755, when only their tonnage is known. However, this was usually cited as 499 tons, which, as Koninckx has suggested, ‘engenders the suspicion that the English company had an interest holding the tonnage under the 500 mark’, suggesting that by undervaluing the tonnage taxes could be dodged. 46 For the periods 1770–1775 and 1783–1792, more realistic figures are available.
Table 5 shows that – apart from the largest Swedish ship – the VOC usually had the ships with the smallest length/breadth ratio, implying that, in relative terms, these ships were the widest ships in the northwest European-Asian traffic. Koninckx characterized the Swedish ships as ‘lumbering East Indiamen’ and also pointed to the fact that the ships’ officers ‘worried little about the dimensions except possibly for the depth,’ which ‘determined how the vessel had to moor in various roadsteads and whether indeed a port could be visited at all.’ 48 The problem of shallow entrances to ports was apparently not an issue exclusive to the VOC. Indeed, the EIC had to deal with the fact that ‘ships bound for Bengal had to draw no more than 17ft in order to navigate the River Hooghly’. 49 Table 5 shows that most ships – irrespective of their company – seldom went over the 17ft mark. Quite remarkably, despite the notorious shallowness of the Dutch estuaries, the VOC ships of the largest (first) charter had the largest ‘hold’, meaning that these ships drew deeper than the ships of all other companies. This increased their carrying capacity, which meant that, especially in the second half of the eighteenth century, the VOC operated ships with the largest carrying capacity relative to their length. It was made possible because these ships returned either to the deep roadstead of Rammekens (for Zeeland) or to Texel (for Amsterdam, Hoorn or Enkhuizen), from where they could negotiate (partly unloaded) the Zuyderzee on their way to Amsterdam, assisted by camels.
Ships’ characteristics.
Note: Results for the EIC have been calculated as an average of the ships of that size which sailed from China in the period concerned; Gøbel and Koninckx only give characteristics of a small number of ships. 47
To be able to make an in-depth analysis of the ships’ characteristics, we need to have accurate design drawings (and/or accurate models) because length, breadth and hold only give partial information. As Figure 1 shows, within a given breadth, the submerged form of the hull could have varied quite significantly. Of the original Bentam VOC design there exist only some technical drawings for the first charter and there exist two models (made by Bentam), one of a ship of the first charter (complete with camels) and one of the third charter. 50 Koninkcx mentions the existence of models of Swedish East Indiamen and he has two sets of drawings. He gives details on the sailplan of the Riksens Ständer, but not for other ships. 51 Gøbel makes no reference to models, drawings or sailplans for the Danish ships, but Sutton shows a construction plan and a sailplan for the Danish Dannemark of 1781, adding that no such sailplans for EIC ships of that period have been found. The only drawings of EIC ships in Sutton’s work are of ships dating from the nineteenth century. 52 Farrington makes no reference to either drawings or sailplans.
What can be ascertained is the difference that the deployment of ‘three-deckers’ might have made. For the VOC, we know which ships were three-decked ships and which were not. 53 Comparing the sailing times of these groups of ships on the return from China gives us an impression of the assumed advantage that three-deckers could master stormy conditions better that ships of the classic design and that they were faster. The numbers in Table 6 show some advantage for the three-deckers over the classic design in respect of the average sailing times. However, the median for both sections of the route are almost exactly the same. This suggests that three-deckers were slightly faster, and probably more reliable and predictable, when making the voyage from Canton to the Republic.
Three-deckers and the classic design compared.
Conclusion
This article examines the eighteenth-century voyages between Asia and Europe in which speed was essential, in particular the homebound passage from Canton in China with tea for European markets. In contrast to the widely held assumption that the ships of the VOC were slower than those of their competitors, a comparison between Danish, English, Swedish and Dutch ships returning from Canton shows that the Dutch could compete very well with their competitors, even to the extent that by 1800 they were closing the gap with the fast Danish ships.
Looking at a number of possible explanations, it becomes clear that choosing the right time of arrival in Canton (for which the VOC had a useful system in place), leading to a suitable month of departure, was an important factor. The experience of skippers also paid off, although one gets the impression that the VOC could have taken more advantage of it. The fact that the EIC was some years ahead of the VOC and the Dansk Asiatisk Kompagni in coppering its ships did not result in a faster voyage for the English ships between Canton and the southern tip of Africa, although the voyage from there to Europe was indeed faster in the closing decades of the eighteenth century. The most likely explanation for that advantage is that the English had developed a better knowledge and understanding of the wind, weather and currents in the South Atlantic, enabling them to steer a faster course. Even when the skippers of the VOC had access to more accurate information, they still had the obligation to stick to the detailed sailing instructions laid down by the VOC that allowed no deviation from the established route of old.
This article further indicates that even when an adaptation of VOC ships between 1714 and 1742 to shallow estuaries in the Republic may have been an explanation for an initial gap in speed, that adaptation had disappeared by the mid-1740s and therefore the initial gap probably disappeared. Another adaptation to the problem of shallow estuaries took place in the second half of the eighteenth century (although only by the Delft and Rotterdam yards, and therefore in small numbers) and cannot therefore be invoked to explain a gap that presumably already existed in the first half of that century. In that respect, looking at the characteristics of ships threw up a surprise. Despite the notorious shallow estuaries of the Dutch delta, the largest ships (the so-called ‘first charter’) of the VOC were the ships that treaded the most deep of all companies concerned. That was made possible by the deep sea entrance to Rammekens (Zeeland), the possibility to sail a (partially laden) ship from the roadstead of Texel to the ports of Hoorn or Enkhuizen and to use camels to ferry (partially laden) ships to Amsterdam. The ports of Delft and Rotterdam could only be reached by the smaller ships of the second charter via the roadstead of Goeree. It seems that the conclusion by Solar and De Zwart that there was an initial gap, widening in later years, between the speed of VOC ships and (in particular) EIC ships during the late eighteenth century cannot be upheld if the focus is on the voyages from China.
Finally, to give due to the staff of the VOC in Canton, when they argued (in 1788) for the exclusive use of (coppered) three-deckers on the China route, they were in all likelihood right. Looking at the sailing times by the three-deckers, one gets the impression that these ships (developed in Zeeland, with the same hull as the classic Bentam ships) were more reliable in terms of the average time it took them to sail home from China. Having them coppered might have increased their speed. The staff in Canton also contended that the skippers be given up-to-date foreign charts and a free hand in choosing the most suitable route. If experienced skippers on the China route had been given more freedom to navigate coppered three-deckers, using the foreign charts finally provided to the China ships, the myth of the slowness of the ships of the VOC might have been punctured once and for all.
Footnotes
Acknowledgements
I wish to thank Pim de Zwart for the meeting during which we discussed the article by Solar and De Zwart, for his reworking of the DAS database (
) which contains all the Dutch-Asiatic shipping movements of the VOC, for his searchable version of Anthony Farrington Catalogue of East India Company Ships’ Journals and Logs 1600-1834 (London, 1999) and for the access he provided to a number of articles. His comments on earlier drafts of this article have been very constructive and helpful. Thanks are also due to Adri Albert de la Bruhèze for his advice during the early stages of the writing of this article.
Author’s note
From 15 October 2020 onwards, the author can be contacted at the following corresponding address: J.C.J. van Speijkstraat 31, 7772 ZB Hardenberg, The Netherlands.