Shipbuilding underwent a virtual revolution in the fifteenth century, although the vital adoption of the stern rudder was at least two centuries earlier. This revolution sprang from a very fruitful interaction between the traditions of the Mediterranean and the northern seas. Iberian builders, particularly those of Biscay, played an important role in this development, without which the Discoveries would not have been possible.^[62]

Mediterranean oared craft ranged from light fighting galleys to the great Venetians which sailed to Southampton and Bruges: three-masted and lateen-rigged, using their banks of oars in calms and for entering and leaving port, they were not only the largest but also the most dependable ships of their day—- 16 -at any rate outside the China Seas!—and hence favoured for passengers and valuable cargoes. Minor cargo in the coastal trade was largely carried by small or medium-sized lateen-rigged craft, fast and readily manoeuvrable, especially in light airs. Keels were curved, and hence a vessel could not be simply beached or allowed to settle aground on the ebb, but had to be shored up; this was not so serious a handicap in the almost tideless Mediterranean as in the Atlantic, though it made careening difficult. A major limitation to the usefulness of both oared and lateen-rigged craft, on long voyages, was their demand for large crews, with consequent reduction in pay-load. On the other hand, the Mediterranean carvel-built construction, with planking edge to edge and bolted or pegged to stout ribs, was superior to the over-lapping clinker-built sides of the northern ships, and in the sixteenth century clinker building was abandoned except for small coastal craft and sometimes for upper works.

The standard northern merchantman was the cog, roomier than equivalent Mediterranean vessels, better fitted to rough seas, and, with its straight keel, able to ground on the ebb without damage. In the fourteenth century the cog had usually only one mast carrying a single large square sail, but improvements in rigging and the handling of sails were continuous. Square-rig called for fewer seamen on large ships—for a vessel of 250 tons, say twenty men, as against fifty for lateen. Hence for bulk traffic where speed was not a primary factor, such as the very important alum trade, square-rig became generally adopted in the Mediterranean except for small coasting and fishing craft, and its advantages over lateen in the heavier weather of the open Atlantic were soon recognised: square sails were much easier to handle and to furl in strong winds than lateen on their very long yards. But the North in turn soon realised the advantage of having more than one mast: either a foremast or a lateen mizzen greatly increased manoeuvrability. Thus from about 1430 a bewildering variety of hybrids were developed, initially it seems largely by the Basques; the technical differences are of intense interest to the cognoscenti.^[63] The end result, the standard big ship for most of the sixteenth century, was the carrack: three masts, with a lateen mizzen, high castles (especially aft), and a large central cargo hatch. This was the nao of the Spanish Carrera and the nau of the Portuguese Carreira to the West and East Indies respectively. By the 1590s such ships sometimes exceeded 1500 tons, though 700 to 1000 would be more usual; these figures had been exceeded by Chinese vessels two centuries earlier, and it is salutary to reflect that three of the essentials for oceanic navigation by large ships—the mariner's compass, multiple masting, the axial rudder—existed in China long before their adoption in Europe. Although their European initiation and development are probably independent, the remarkably rapid flowering of European ship design from about 1450 may owe something to borrowings from China via the Arabs of the Indian Ocean.^[64]

Early in the sixteenth century there was a rash of competitive prestige building of ‘Ships Royal’ such as the Henry Grace à Dieu, which might have four masts with three fighting tops on the main, and fantastic sail plans. More to the point was- 17 -

the invention about 1501–2 of the gun-port, which meant that much heavier armament could be carried, guns to damage masts and hulls instead of light essentially anti-personnel weapons mounted on high ‘castles’ at bows and stern.^[65] The future was not with the huge parade ships but with the galleon, developed as a specialised fighting ship, with lower castles (especially the forecastle) and finer lines than the carrack. Usually between 250 and 500 tons and carrying up to forty guns, some reached 800 or 1000 tons by the end of the century (Figure Plate II, “AN ELIZABETHAN GALLEON. ”).^[66] Galleons formed the escorts of the Spanish trading fleets to America, carrying no licit cargo themselves except royal bullion, although the enormously important Manila-Acapulco run was worked by ‘the Galleon’. Another major evolution of the later sixteenth century was the development of more effective sail plans, including topsails on all masts except the mizzen, which long retained its lateen, and even topgallants.

- 18 -- 19 -

Alongside these greater ships there was of course a host of smaller types, of which the most important was the caravel (Figure Plate III, “CARAVELS”), the main instrument of Portuguese exploration until the Cape had been rounded. The Portuguese caravel—the Spanish version differed—was apparently a home-grown product, developed from small coastal barcas. At first they were very small, under 50 tonéis, only partly decked, with two or three lateen masts; later they reached 150 to 250 tonéis or more, with three or four masts, the caravela redonda having one or two square sails on the foremast. Light and very handy, good at sailing near the wind, they were regarded as very versatile, as indeed they were; but, except for the largest, they had only a very modest superstructure on the poop and provided very little accommodation. Admirable for inshore work, they were not really tough enough for long-distance exploration in the open ocean; but as auxiliaries in war and trade, especially in littoral seas, they lasted until near the end of the seventeenth century.^[67] In the north, the Dutch fluyt or flyboat became prominent as a medium-sized cargo vessel before 1600, and as the felibote played an important part in the colonial trade of Spanish America (Figure Plate IV, “DUTCH FLUYTS. ”).

Finally there was a large assortment of small craft: pinnaces, pataches, barcos, bergantins. This last type must be distinguished from the later brigantines, as is sufficiently shown by the fact that Cortes built thirteen of them in seven weeks

- 20 -for his final attack across the Lake of Mexico.^[68] The original bergantin seems to have been essentially a light galley with an auxiliary lateen (later square) sail, suitable for river or inshore coastal work, the oars making it possible to work against wind or current. Later the term recurs constantly in the records of the Spanish American coastal trade, along with the patache, which was like a small brigantine in the modern sense. In this context also the Spanish fregata, until quite late in the eighteenth century, often meant not a fairly large warship but a small or medium-sized coastal trader or felibote, often built in American yards, for coasting in the first place but capable in emergency of making trans-Atlantic passages.^[69] The maid-of-all-work on English voyages was the pinnace, the counterpart of the bergantin.

It is a far cry from the galley and the cog to the great and complex ships of the seventeenth and eighteenth centuries, with broadside armaments which, unless in extraordinary circumstances, rendered them virtually impregnable to any opponents to be met with in extra-European waters. The continual increase in the size of long-distance ships had a sound economic basis, once coastal or littoral exploration had yielded to exploitation. For initial voyages over long distances out of sight of land, with no known ports of supply and little possibility of estimating the length of the voyage with much accuracy, safety demanded ample provisioning, with consequent loss of cargo space, though economy might dictate the use of smaller vessels, in twos or threes to spread the risk of loss.

To equip a ship of 65 tons for two years' exploring practically ruled out any pay-load, at least on the outward journey, though commodities as valuable for their bulk as bullion or spices could be brought back. But once commerce was established, a ship of 700 tons was much more economic than one of 300; the larger ship, with a crew of eighty or ninety, would demand a ‘poids moteur’—food, stores, wine, water—of only 10 per cent of its transport capacity; the fifty or sixty men on the smaller would need 13 to 15 per cent. Hence the tendency to ever-increasing size in the Carreira da India (and to a lesser degree on the Carrera) and in the great Indiamen of the seventeenth and eighteenth centuries.^[70] One possible limitation on this growth, the need to make up full cargoes, could be met by feeders to a few great ports, the ‘country trade’ of South and Southeast Asia in East India Company days, the cabotaje of the Caribbean and the Pacific in the days of the Puerto Bello fairs. For such miscellaneous carriage, as distinct from the great main lines, fluyts, later on brigs and barques, were essential. The real limitation in size was simply the cost and time of building the giants.

Navigation at the end of the fourteenth century, at least outside the Mediterranean, was still almost entirely a matter of empirical experience, with the simplest instrumentation: little beyond the compass and the lead, greased to bring up samples of the sea-bottom; pilotage rather than true navigation. The range of expertise of the ordinary skipper was still that of Chaucer's Shipman, relying- 21 -on an intensely detailed memory of tides and currents, ports and landmarks, essentially rule of thumb; although could he read, written pilot-books or rutters were probably already available. But the next century saw the gradual, and very uneven, introduction of theory. In the long run this was to change a craft ‘mystery’ into an applied science; already in the early sixteenth century the Casa de Contratacion in Seville had charge of a formal system of examinations for certificating Spanish pilots, who by this time had to acquire some mathematical skills.^[71]

The date at which astronomical navigation was introduced is uncertain, although there can be no doubt that the necessities of their Atlantic voyaging made the Portuguese the pioneers: mere dead reckoning, even assisted by the traverse table, was no longer adequate, as it was in the Mediterranean. The first references to the observation of the altitude of the Pole Star are those by Cadamosto, who was on two voyages with the Portuguese in the 1450s; his terms—‘the height of a lance’ or of a man—do not imply instrumental navigation, but Portuguese students find it difficult to envisage regular two-way voyages between Lisbon and the Azores (officially colonised in 1439) without some techniques for taking heights of the Pole Star.^[72] Be that as it may, before 1480 the astrolabe and the quadrant had been adapted for use at sea (possibly by Prince Henry's Jewish expert, Master Jacome of Majorca), and tables of latitude had been drawn up for points as far south as the Equator—using the sun, for the Pole Star was too low to be easily observed as far south as Guinea. These tables are found in ‘the oldest surviving navigational manual’, the Regimento do Astrolabio e do Quadrante, of which an edition, probably not the first, was printed in Lisbon in 1509.^[73] Later the Jacob's Staff and the back-staff (which avoided direct sights at the sun) superseded the cumbersome astrolabe, until the introduction of improved quadrants by Davis in the seventeenth century and James Hadley in 1731.

So much for latitudes: those of the Regimento are often correct to within ten minutes, so this was no longer a serious problem. The accurate determination of longitude at sea, however, remained in practice impossible for two and a half centuries after Pope Alexander VI had made it ‘a live issue’ by decreeing a meridian as the demarcation line between Portuguese and Spanish hemispheres. The theory was there—Vespucci and the Dieppois Jean Rotz had attempted to use lunar distances before 1540, Columbus tried the timing of a lunar eclipse, and Rotz and others thought that magnetic variation was or might be sufficiently regular in its distribution to give an indication of longitude.^[74] But neither the observational nor the timekeeping instruments available were adequate to attain the precise readings which were needed. In effect, the mariner had to fall back on course steered and distance made. Distance was checked, all too roughly, by various log devices, all crude—though once again theory, with the concept of a geared instrument, was ahead of practicability. As for course steered, the traverse board, on which the time run on each bearing during a given period could be recorded, was an ingenious, if rough and ready, graphical solution. With- 22 -all these devices, latitude sailing—going north or south until one reached the latitude of the destination, then easting or westing—became a practicable and much-used procedure; to be safe from piling up on shore, one usually adopted the biggest possible estimate of longitudinal distance made.^[75] But longitudes still remained a matter of dead reckoning, and the results can be seen in the vagabond habit of Mendaña's Islas de Salomon, discovered in 1568 and in the next two centuries placed anywhere between the longitudes of Cooktown in Queensland, 145°E, and the Marquesas, 140°W—a difference of 75 degrees! Those seen by Mendaña actually lie around 160°E.^[76]

Mediterranean sailors had long had the assistance of the portulan chart, which, especially in its fully developed Catalan form, gave an accurate delineation of the shores of the Mediterranean, and of the Atlantic as far north as the Narrow Seas; it had a linear scale and a system of wind-roses from which a pilot could work out his bearing from port to port.^[77] It had no grid of latitude and longitude, and hence no projection; the earth was treated as a plane surface. For the Mediterranean, with its short north-south span, this did not greatly matter, since the convergence of the meridians over some 15 degrees of middle latitudes was too slight to induce really serious distortion. It was otherwise when the range of latitude involved stretched to the Equator and beyond, still more when the globe itself had to be plotted on a flat sheet.

This problem was not, of course, anything new, but it had hitherto been an academic one; the theory was well within the grasp of Renaissance mathematicians. It was once more the necessities of Portuguese navigation in the Atlantic which led to the first steps, again seemingly at the hands of the learned Master Jacome. Initially, these steps were modest enough, merely the addition to portulan-type charts of a north-south line marked off in degrees of latitude, originally just a line on magnetic north, later allowing for the variation by a similarly divided true meridian at the appropriate acute angle. Tables of the length of a degree of longitude according to latitude were produced, and early in the sixteenth century the Portuguese Jew Pedro Nuñes devised a quadrant by which these values could be read off directly. He also worked out the true spiral form of rhumb-lines—lines to intersect all meridians at a constant angle—but this was well above the heads of practical seamen, who needed a simple chart on which such a course of constant bearing could be plotted as a straight line. He was unable to provide this, but he led the way to Mercator—or perhaps more correctly Edward Wright—who did.^[78] Nuñes anticipated Jonathan Swift in a fine, but more scientific, scorn for cartographers who used plenty of gold paint and planted all over the place flags, camels, and ‘elephants for want of towns.’