Early medieval Europeans received from their predecessors two broad ranges of wooden shipbuilding traditions, one in the Mediterranean and the other in the northern seas. At the same time Chinese shipwrights had already developed the central features of the design of the junk. Its watertight compartments, adjustable keel, and highly flexible number of masts each carrying a lug sail with battens, made the junk a highly versatile and reliable seagoing ship. Junks by the year 1000 were much larger than any ships in Europe or in the great oceanic area where a Malaysian shipbuilding tradition predominated. There, ocean-going rafts with outriggers or twin hulls and rigged with, at first, bipole masts ranged much more widely than vessels from any other part of the world carrying the designs and building practices across the Indian Ocean to Madagascar and around the Pacific Ocean to the islands of Polynesia. Along the shores of the Arabian Sea shipbuilders constructed dhows, relatively shallow cargo vessels rigged with a single triangular or lateen sail. The planks of the hulls were typically sewn together with pieces of rope, a loose system which made the hull flexible, and so able to handle rough seas, but not very watertight. There were also serious limitations on how big such hulls could be built, unlike junks where vessels of one thousand tons and more seem to have been feasible.

Mediterranean Practice

Roman shipbuilders followed Greek practices in building their hulls with mortise and tenon joints. Wedges or tenons were placed in cavities or mortises gouged out of the planks and held in place by wooden nails passed through the hull planks and the tenons. In the Roman Empire the methods of fastening predominated on all parts of ships, including the decks, and the tenons were very close to each other. The resulting hull was extremely strong, heavy, and sturdy so the internal framing was minimal. The hull was also very watertight but even so the surface was often covered with wax or even copper sheathing to protect it from attack by shipworm (Teredo navalis). Propulsion came from a single square sail stepped near the middle of the ship. Often the mainsail was supplemented with a small square sail slung under the bowsprit. Roman shipbuilders produced vessels of two general categories, round ships with length-to-breadth ratios of about 3:1 propelled entirely by sails, and galleys with length-to-breadth ratios of about 5:1 propelled both by the standard rig and by oars. Although it was possible to have multiple banks of rowers, in the Roman Empire there was typically only one, with each rower handling a single oar. Shipbuilders gave all those vessels at least one but often two side rudders for control.

As the economy declined in the early Middle Ages and the supply of skilled labor was reduced, the quality of shipbuilding deteriorated. The distance between mortise and tenon joints increased, and on the upper parts of hulls such joints disappeared entirely with planks merely pinned to internal frames. The trend led by the end of the first millennium C.E. to a new form of hull construction. Instead of relying on the exterior hull for strength, shipbuilders transferred the task of maintaining the integrity of the vessel to the internal frame. The process of ship construction as a result reversed, with the internal ribs set up first and then the hull planks added. The planks were still fitted end-to-end as with the old method but now to maintain watertightness they needed to be caulked more extensively and more regularly. The internal frames gave shape to the hull so their design became much more important. The designer of those frames in turn took on a significantly higher status, the hewers of the planks a lesser position. The new type of skeleton-first construction made for a lighter and more flexible ship which was easier to build, needed less wood, but required more maintenance. Increasing the scale of the ship or changing the shape of the hull was now easier. Builders used the new kind of construction both on large sailing round ships and oared galleys.

In the course of the early Middle Ages Mediterranean vessels went through a change in rigging as well. Triangular lateen sails were in use in classical Greece and Rome for small vessels. As big ships disappeared with the decline of the Roman Empire and economy the lateen sail came to dominate and square sails all but disappeared. Lateen sails had the advantage of making it possible to sail closer to the wind. Lateen sails had the disadvantage that when coming about, that is changing course by something of the order of ninety degrees, the yard from which the sail was hung had to be moved to the other side of the mast. In order to do that the yard had to be carried over the top of the mast, which was a clumsy, complex, and manpower-hungry operation. There was a limitation then on the size of sails and thus on the size of ships. It was possible to add a second mast, which shipbuilders often did both on galleys and on round ships since that was the only way to increase total sail area.

Northern European Practice

Shipbuilders around the Baltic and North Seas in the early Middle Ages produced a variety of different types of vessels which were the ancestors of a range of craft that melded together over the years to create one principal kind of sailing ship. The rowing barge was a simple vessel with overlapping planking. The planks could be held fast by ropes but over time shipbuilders turned to wooden nails or iron rivets for the purpose. That type of lapstrake construction for hulls meant that internal ribs were of little importance in strengthening the hull. At first shipwrights used long planks running from bow to stern but they discovered that by scarfing shorter pieces together not only did they eliminate a constraint on the length of their vessels but they also increased the flexibility of the hulls. At some point, probably in the eighth century, the rowing barge got a real keel and also a single square sail on a single mast stepped in the middle of the ship. The new type, with both ends looking much the same, was an effective open ocean sailor. Scandinavian shipbuilders produced broadly two versions of what can be called the Viking ship after its most famous users. One version was low, and fitted with oars and a mast that could be taken down or put up quickly and with a length-to-breadth ratio of 5:1 or 6:1. The other version had a fixed mast, few if any oars at the bow and stern which were there just to help in difficult circumstances, and a length-to-breadth ratio of around 3:1. Both types had a single side rudder which apparently gave a high degree of control. The Viking ship evolved into a versatile cargo ship which was also effective as a military transport and warship. Often called a keel because of one of the features which allowed it to take to the open ocean, it was produced in variations throughout northern Europe and along the Atlantic front as far south as Iberia.

The other types that came from early medieval northern shipyards were more limited in size and complexity. The hulk had a very simple system of planking which gave way over time to lapstrake construction. The hull had the form of a banana and there was no keel so it proved effective in use on rivers and in estuaries. The hull planks, because of the shape of the hull, met at the bow in a unique way and were often held in place by tying them together. Rigging was a single square sail on a single mast which could be, in the case of vessels designed for river travel, set well forward. The cog had a very different form from the hulk. While the planks on the sides overlapped there was a sharp angle between those side planks and the ones on the bottom. Those bottom planks were placed end-to-end and the floor was virtually flat. With posts at either end almost vertical the hull was somewhat box-like. The type was suited to use on tidal flats where it could rest squarely on the bottom when the tide was out, be unloaded and loaded, and then float off when the tide came in. There was a single square sail on a single mast placed in the middle of the ship. The design certainly had Celtic origins but it was transformed by shipwrights in the High Middle Ages to make it into the dominant cargo and military vessel of the North.

Shipbuilders, possibly in the Low Countries, gave the cog a keel. In doing that they also made changes in the form of the hull, overlapping the bottom planks and modifying the sharp angles between the bottom and side planks. The result was a still box-like hull which had greater carrying capacity per unit length than keels. The cog could also be built higher than its predecessors but that meant passing heavy squared timbers through from one side to the other high in the ship to keep the sides in place. Shipbuilders fitted the hull planks into the heavy posts at the bow and stern and also fixed a rudder to the sternpost which was more stable than a side rudder. In the long run it would prove more efficient as well. Cogs could be and were made much larger than other contemporary vessels. Greater size meant a need for a larger sail and a larger crew to raise it. To get more sail area sailors added a bonnet, an extra rectangular piece of canvas that could be temporarily sewn to the bottom of the sail. That gave the mariners greater flexibility in deploying canvas without increasing manning requirements. Riding higher in the water and able to carry larger numbers of men than other contemporary types cogs became the standard vessels of northern naval forces, doubling as cargo ships in peacetime.

While the two shipbuilding traditions of the Mediterranean and northern Europe remained largely isolated through the early and High Middle Ages, from the late thirteenth century both benefited from extensive contact and borrowing of designs and building methods. Sailors in southern Europe used the cog certainly by the beginning of the fourteenth century and probably earlier. Shipwrights in the Mediterranean appreciated the advantages of greater carrying capacity but they were also conscious of the limitations set by the simple rig. They added a second mast near the stern and fitted it with a lateen sail. They also changed the form of hull construction, going over to skeleton-first building. The result was the carrack, in use by the late fourteenth century. It was easier to build, probably lighter than a cog of the same size, and could be built bigger. Most of all the two masts and the presence of a triangular sail gave mariners greater control over their vessels and made it possible for them to sail closer to the wind. The next logical step, taken sometime around the end of the fourteenth century, was to add a third small mast near the bow to balance the one at the stern. The driving sail and principal source of propulsion was still the mainsail on the mainmast but the combination or full-rig made ships more maneuverable and able to sail in a greater variety of conditions. While older forms of ships, such as the keel or the cog or the lateen-rigged cargo ship of the Mediterranean, did not by any means disappear, the full-rigged ship came to dominate exchange over longer distances, especially in the form of the full-rigged carrack travelling between southern and northern Europe. Northern Europeans were slow to adapt to skeleton-first hull construction, in some cases even combining old methods with the new one. By the end of the fifteenth century the full-rigged ship was the preferred vessel for many intra-European trades, in part because of its handling qualities, in part because of its versatility, and in part because its crew size could be reduced per ton of goods carried compared to other types. The greater range also led to its replacing, for example, the simpler, lower, lateen-rigged caravel in Portuguese voyages of exploration along the west coast of Africa. Full-rigged ships in daily use were the choice for voyages of exploration and became in the Renaissance the vehicles for European domination of the ocean seas and for the resulting international trading connections and colonization.


The Arabian Peninsula, surrounded by the Red Sea, the Indian Ocean, and the Arabian (or Persian) Gulf, had a geostrategic position in relations between East and West. Before the beginning of Islam in 622 the Arabs had had some nautical experience which was reflected in the Qur’an (VI, 97: “It is He who created for you the stars, so that they may guide you in the darkness of land and sea”; XIV, 32: “He drives the ships which by His leave sail the ocean in your service”; XVI, 14: “It is He who has subjected to you the ocean so that you may eat of its fresh fish and you bring up from it ornaments with which to adorn your persons. Behold the ships plowing their course through it.”) and in ancient poetry (some verses of the poets Tarafa, al-A‘sha’, ‘Amr Ibn Kulthum, and others). Arabs used the sea for transporting goods from or to the next coasts and for the exploitation of its resources (fish, pearls, and coral). However, their experience in maritime matters was limited due to the very rugged coastline of Arabia with its many reefs and was limited to people living on the coast. Because they lacked iron, Arab shipwrights did not use nails, but rather secured the timbers with string made from palm tree thread, caulked them with oakum from palm trees, and covered them with shark fat. This system provided the ships with the necessary flexibility to avoid the numerous reefs. The Andalusi Ibn Jubayr and the Magribi Ibn Battuta confirm these practices in the accounts of their travels that brought them to this area in the thirteenth and fourteenth centuries, respectively. Ibn Battuta also notes that in the Red Sea people used to sail only from sunrise to sunset and by night they brought the ships ashore because of the reefs. The captain, called the rubban, always stood at the bow to warn the helmsman of reefs.

The regularity of the trade winds, as well as the eastward expansion of Islam, brought the Arabs into the commercial world of the Indian Ocean, an experience that was reflected in a genre of literature which mixes reality and fantasy. Typical are the stories found in the Akhbar al-Sin wa-l-Hind (News of China and India) and ’Aja’ib al-Hind (Wonders of India), as well as the tales of Sinbad the Sailor from the popular One Thousand and One Nights. But medieval navigation was also reflected in the works of the pilots such as *Ahmad Ibn Majid (whose book on navigation has been translated into English) and Sulayman al-Mahri.

In the Mediterranean Sea

The conquests by the Arabs of Syria and Egypt in the seventh century gave them access to the Mediterranean, which they called Bahr al-Rum (Byzantine Sea) or Bahr al-Sham (Syrian Sea). Nautical conditions were very different in this sea: irregular but moderate winds, no heavy swells, and a mountainous coastline that provided ample visual guides for the sailors on days with good visibility. The Arabs took advantage of the pre-existing nautical traditions of the Mediterranean peoples they defeated. In addition, we have evidence for the migration of Persian craftsmen to the Syrian coast to work in ship building, just as, later on, some Egyptian craftsmen worked in Tunisian shipyards.

The Arab conquests of the Iberian Peninsula (al-Andalus) and islands such as Sicily, Crete, and Cyprus set off a struggle between Christian and Muslim powers for control of the Mediterranean for trade, travel, and communications in general. Different Arab states exercised naval domination of the Mediterranean, especially during the tenth century. According to the historian Ibn Khaldun, warships were commanded by a qa’id, who was in charge of military matters, armaments, and soldiers, and a technical chief, the ra’is, responsible for purely naval tasks. As the Arabs developed commercial traffic in Mediterranean waters, they developed a body of maritime law which was codified in the Kitab Akriyat al-sufun (The Book of Chartering Ships). From the end of the tenth century and throughout the eleventh, Muslim naval power gradually began to lose its superiority.

In navigation technique, the compass reached al-Andalus by the eleventh century, permitting mariners to chart courses with directions added to the distances of the ancient voyages. The next step was the drawing of navigational charts which were common by the end of the thirteenth century. Ibn Khaldun states that the Mediterranean coasts were drawn on sheets called kunbas, used by the sailors as guides because the winds and the routes were indicated on them.

In the Atlantic Ocean

The Atlantic coasts of Europe and Africa, despite their marginal situation with respect to the known world at that time, had an active maritime life. The Arabs usually called this Ocean al-Bahr al-Muhit (“the Encircling or Surrounding Sea”), sometimes al-Bahr al-Azam (“the Biggest Sea”), al-Bahr al-Akhdar (“the Green Sea”) or al-Bahr al-Garbi (“the Western Sea”) and at other times al-Bahr al-Muzlim (“the Gloomy Sea”) or Bahr al-Zulumat (“Sea of Darkness”), because of its numerous banks and its propensity for fog and storms. Few sailors navigated in the open Atlantic, preferring to sail without losing sight of the coast. The geographer *al-Idrisi in the middle of the twelfth century informs us so: “Nobody knows what there is in that sea, nor can ascertain it, because of the difficulties that deep fogs, the height of the waves, the frequent storms, the innumerable monsters that dwell there, and strong winds offer to navigation. In this sea, however, there are many islands, both peopled and uninhabited. No mariners dare sail the high seas; they limit themselves to coasting, always in sight of land.” Other geographers, including Yaqut and al-Himyari, mention this short-haul, cabotage style of navigation. Yaqut observes that, on the other side of the world, in the faraway lands of China, people did not sail across the sea either. And al-Himyari specifies that the Atlantic coasts are sailed from the “country of the black people” north to Brittany. In the fourteenth century, Ibn Khaldun attributed the reluctance of sailors to penetrate the Ocean to the inexistence of nautical charts with indications of the winds and their directions that could be used to guide pilots, as Mediterranean charts did. Nevertheless, Arab authors describe some maritime adventurers who did embark on voyages of exploration.

Fluvial Navigation

Only on the great rivers such as the Tigris, the Euphrates, the Nile and, in the West, the Guadalquivir was there significant navigation. It was common to establish ports in the estuaries of rivers to make use of the banks to protect the ships.

Nautical Innovations

Two important innovations used by the Arabs in medieval period are worthy of mention: the triangular lateen sail (also called staysail), and the sternpost rudder. The lateen sail made it possible to sail into the wind and was widely adopted in the Mediterranean Sea, in view of its irregular winds. The Eastern geographer Ibn Hawqal, in the tenth century, described seeing vessels in the Nile River that were sailing in opposite directions even though they were propelled by the same wind. To mount only one rudder in the sternpost which could be operated only by one person proved vastly more efficient that the two traditional lateral oars it replaced. Although some researchers assert that this type of rudder originated in Scandinavia and then diffused to the Mediterranean Sea, eventually reaching the Arabs, it is most likely a Chinese invention which, thanks to the Arabs, reached the Mediterranean.

Toward Astronomical Navigation

The Arabs made great strides in astronomical navigation in the medieval period. With the help of astronomical tables and calendars, Arab sailors could ascertain solar longitude at a given moment and, after calculating the Sun’s altitude as it comes through the Meridian with an astrolabe or a simple quadrant, they could know the latitude of the place they were in. At night, they navigated by the altitude of the Pole Star. For this operation Arab sailors in the Indian Ocean used a simple wooden block with a knotted string called the kamal which was used to take celestial altitudes. They also knew how to correct Pole Star observations to find the true North. Ibn Majid, for example, made this correction with the help of the constellation called Farqadan, which can only be seen in equatorial seas.

The determination of the longitude was a problem without a practical solution until the invention of the chronometer in the eighteenth century. Arab sailors probably may have used a sand clock to measure time, because they knew how to produce a type of glass that was not affected by weather conditions. So they could estimate the distance that the ship had already covered, even though speed could not be accurately determined. The navigational time unit used was called majra, which the geographer Abu l-Fida’ defines as “the distance that the ship covers in a day and a night with a following wind,” a nautical day that it is the rough equivalent of one hundred miles.


Bass, George, ed. A History of Seafaring Based on Underwater Archaeology. London: Thames and Hudson, 1972.

Friel, Ian. The Good Ship: Ships, Shipbuilding and Technology in England, 1200–1520. London: British Museum Press, 1995.

Gardiner, Robert, ed. The Earliest Ships: the Evolution of Boats into Ships. London: Conway Maritime Press, 1996.

——, ed. Cogs, Caravels and Galleons The Sailing Ship 1000–1650. London: Conway Maritime Press, 1994.

Hattendorf, John B., ed. Maritime History in the Age of Discovery: An Introduction. Malabar, Florida: Krieger, 1995.

Hutchinson, Gillian. Medieval Ships and Shipping. Rutherford: Fairleigh Dickinson University Press, 1994.

Lane, Frederic C. Venetian Ships and Shipbuilders of the Renaissance. Baltimore: Johns Hopkins Press, 1934.

Lewis, Archibald R. and Timothy J. Runyan. European Naval and Maritime History, 300–1500. Bloomington: Indiana University Press, 1985.

McGrail, Seán. Boats of the World from the Stone Age to Medieval Times. Oxford: Oxford University Press, 2001.

Pryor, J. H. Geography, Technology and War: Studies in the Maritime History of the Mediterranean 649–1571. New York: Cambridge University Press, 1988.

Unger, Richard W. The Ship in the Medieval Economy, 600–1600. London: Croom-Helm Ltd., 1980.

Fahmy, Aly Mohamed. Muslim Naval Organisation in the Eastern Mediterranean from the Seventh to the Tenth Century A.D. 2 vols. Cairo: General Egyptian Book Organisation, 1980.

Lewis, Archibald Ross. Naval Power and Trade in the Mediterranean, A. D. 500–1000. Princeton: Princeton University Press, 1951.

Lirola Delgado, Jorge. El poder naval de Al-Andalus en la época del Califato Omeya. Granada: Universidad de Granada, 1993.

Picard, Christophe. La mer et les musulmans d’Occident au Moyen Age. VIIIe–XIIIe siècle. Paris: Presses Universitaires de France, 1997.

Pryor, John H. Geography, Technology and War. Studies in the Maritime History of the Mediterranean, 649–1571. New York: Cambridge University Press, 1988.

Tibbetts, G. R. Arab Navigation in the Indian Ocean before the Coming of the Portuguese being a translation of Kitab al-Fawa’id fi usul al-bahr wa’l-qawa’id of Ahmad b. Majid al-Najdi. London: Royal Asiatic Society, 1971.


HMS Inflexible (1876)

Inflexible, 1876, as completed with sails for training. Note the torpedo launching chute over the stem.

The design concept of Inflexible was of a raft, the citadel, which would float if the ends were destroyed or flooded. The ends were closely subdivided and protected by a thick deck. A light, unprotected structure above provided accommodation.

In 1885 Inflexible’s sailing rig was replaced by two military masts.

In a letter to The Times of 1 January 1877, Edward Reed described the Inflexible as `… a huge engine of war, animated and put into activity in every part by steam and steam alone. The main propelling engines are worked by steam, a separate steam engine starts and stops them; steam ventilates the monster, steam weighs the anchors, steam steers her, steam pumps her out if she leaks, steam loads the gun, steam trains it, steam elevates or depresses it. The Ship is a steam being .’

The 1873 Estimates envisaged the building of a single, improved ‘Fury’ (in fact, this meant Fury, not yet renamed, with the modifications which made her Dreadnought). The problem facing Barnaby was stark; the 12.5in, 38-ton gun fitted in recent ships could fire an 820lb projectile through 15.7in of iron armour at 1000yds. Fury’s 14in belt (amidships) was already inadequate and, furthermore, both Woolwich and Elswick claimed that 50-ton guns were within existing capabilities with even larger guns in the near future.

The early studies retained the main features of Dreadnought with the two twin 38-ton turrets augmented by a number of smaller guns en barbette amidships. In one such study a single 50-ton gun in a turret was squeezed in amidships. The 14in belt was retained amidships but the thinner belt at the ends was omitted and a thick transverse bulkhead fitted at each end of the belt. Thus the much admired end-to-end belt of Devastation was already abandoned for what must have been a very small saving in weight.

By this time Woolwich was speaking with confidence of a 60-ton gun and Barnaby was driven to a more radical solution. The main requirements seem to have been set by Barnaby himself, though presumably after discussion with Board members and others. The armament was to consist of two twin turrets with 60-ton guns capable, if possible of being changed to 80-ton guns when available. White described the problem: ‘At first it was contemplated to have 60-ton guns and the ship was laid down on this basis. Finally, in 1874 it was decided to adopt 80-ton guns, which involved an increased weight aloft of 200 tons, and considerably modified the design, the draft and displacement having to be increased. There had been some previous instances of ships getting ahead of the settlement of their gun designs but never so serious one as this. Unfortunately, it was only the first of a long series of similar difficulties … .’ The armour was to be concentrated over a short citadel with a maximum thickness of 24in. She was to be fast – 14kts – and capable of using the Suez Canal at light draught (24ft 4in). Barnaby’s ideas were generally welcomed and the design was progressed incorporating some detail improvements mainly suggested by the DNO, Captain Hood, but with some later ideas from Barnaby. The following paragraphs describe the design as it finally evolved.

The design concept was of a very heavily armoured raft containing the machinery and magazines on which the two turrets were carried. The ends were protected by a strong armoured deck below the waterline, by close subdivision and by buoyant material whilst a light superstructure provided living space. Even if both ends were flooded, the armoured box was intended to have sufficient buoyancy and stability to float upright. This stability requirement led to a wide beam which, in turn, meant that the turrets could fire close to the axis past the narrow superstructure, limited by blast damage to the superstructure. She was fitted with anti-rolling water tanks to reduce the severity of rolling but these were ineffective.

The earliest studies of this configuration showed 60-ton guns though provision was made to mount 100-ton guns when they became available. Woolwich built an experimental 80-ton MLR which completed in September 1875 with a 14.5in bore. After tests, it was bored out to 15in and after further tests in March 1876 it was finally enlarged to 16in bore with an 18in chamber, accepting a 370lb charge. This gun fired a total of 140 rounds-215,855lbs of iron from 42,203lbs of powder – mostly against what was known as ‘Target 41’ which had four 8in plates separated by 5in teak. The standard system of grooving used with studded shell proved troublesome and in final form it had thirty-nine shallow grooves (‘polygroove’) with a lead gas check at the base of the shell.

The production guns-80-ton, Mark I-were mounted in twin turrets each weighing 750 tons and 33ft 10in external diameter. These turrets had an outer layer of compound armour with 18in teak backing and an inner layer of 7in wrought iron. The projectile weighed 16841b and when fired with the full charge of 450lbs brown prism powder had a muzzle velocity of 1590ft/sec and in tests could penetrate 23in of wrought iron in either a single thickness or two plates spaced. The interval between rounds was said to be between 2½ and 4 minutes. To load, the guns were run out and depressed against ports in the deck through which hydraulic rams loaded the guns. Two of these monstrous guns survive on the train ferry pier at Dover, though the turret design is rather different and an early studded shell is in the Naval Armament Museum, Gosport.

Inflexible’s citadel was protected at the waterline by a strake of 12in plate, 4ft deep, backed by 11 in teak containing vertical frames. Behind this was another 12in plate backed by 6in horizontal frames, filled with teak followed by the shell of two thicknesses of ⅝in plate. The total thickness of this waterline belt was 4lin, weighing 1100lbs/sq ft and this thickness was preserved in the protection above and below, the thickness of teak increasing as that of the iron was reduced. Above the waterline strake there was a 12in outer plate and an 8in inner plate whilst below the thicknesses were 12in and 4in.

It is not clear why the armour was in two thicknesses as a 22in plate was made by 1877 and it was already recognised that two plates are inferior to a single plate of the same total thickness. A test in 1877 showed that a single plate 17-17½in thick was equivalent to three plates of 6½in. The waterline belt of 24in in total was the thickest belt ever carried on a battleship but it was only 4ft high and would have been of limited value. It does not seem that this protection was tried in final form. It was claimed that this protection was invulnerable to guns similar to those she carried and even to the 17.7in, 100-ton Elswick guns mounted in Italian ships but it was clearly the end of the road for wrought iron as the weight was already at the very limit of what could be carried.

The protection for the ends was a very sophisticated combination of measures. The first line of defence was a 3in wrought iron deck, normally 6-8ft below the waterline. The space between this deck and the middle deck, just above water, was closely subdivided and used for coal and stores which would limit the amount of water which could enter from holes in the side. In addition, narrow tanks 4ft wide and filled with cork were arranged at the sides between these decks and extending 4ft above the middle deck. Inside these cork-filled spaces there was a 2ft coffer dam filled with canvas packed with oakum. All these fillings were treated with calcium chloride to reduce their flammability although tests showed this was not very effective. This scheme has much in common with that which Reed proposed to the 1871 Committee.

In 1877, Reed wrote to Barnaby and later to The Times claiming that calculations which he and Elgar had made showed that the stability provided by the citadel was inadequate if both ends were flooded. Despite a comprehensive rebuttal by Barnaby, an enquiry was set up chaired by Admiral Hope and consisting of three distinguished engineers, Wooley, Rendel and W Froude. Their investigation was extremely thorough, entering into aspects of naval architecture never previously studied.

Their report concluded that it was most unlikely that both ends would be completely flooded but that if this did happen, the Inflexible would a retain a small but just adequate margin of stability in terms of the GZ curve. Their comments on the difficulty of actually hitting the enemy ship are of interest – remember the Glatton turret and Hotspurs initial miss! They listed the problems as the relative movements of the two ships, the smoke generated (470lbs of powder per round), the rolling and pitching of the firing ship, the lack of any way of determining range and the deflection due to wind. In particular, they noted that it was customary to fire the guns from a rolling ship when the deck appeared horizontal at which position the angular velocity was greatest. (Note also that Froude had showed that human balance organs are very bad at determining true vertical in a rolling ship.) All in all, hits anywhere on the ship would be few and those in a position to flood the ends few indeed.

A shell exploding within the cork would destroy it locally but tests showed that a shell hitting light structure would explode about  of a second later during which it would travel 6-10ft, clear of the cork. The canvas and oakum filling of the coffer dam was quite effective at reducing the size of the hole made by a projectile passing through. Both the cork and the coffer dam were tested full scale with the gunboat Nettle firing a 64pdr shell into replicas. The Committee also pointed out that shells were unlikely to enter the space between the waterline and armoured deck except at long range when hits were even less likely.

Though the Committee thought it was unlikely that the ends would be riddled (filled with water) and even less likely that they would be gutted (all stores, coal, cork etc, blown out with water filling the entire space), they examined these conditions with extreme care. Stability curves were prepared and Froude carried out rolling trials on a 1-ton model both in his experiment tank at Torquay and in waves at sea. The movement of floodwater within the ship acted to oppose rolling in waves, as in an anti-rolling tank. The effect of speed on the trim of the flooded model was also examined. Their conclusion was that the ship should survive this extreme condition but would be incapable of anything other than returning for repair.

This investigation was far more thorough than any previous study of the effects of damage and owed much to White’s calculations and Froude’s experiments. It was the first time that GZ curves of stability had been drawn for a damaged ship and the importance of armoured freeboard was brought out and it must be a matter for regret that similar work was not carried out for later ships. With the invaluable gift of hindsight, one may suggest two aspects not fully brought out. The first was the vulnerability of the citadel armour itself, particularly bearing in mind the shallow 24in layer, in two thicknesses, and the increasing power of guns. The second point was the assumption that the watertight integrity of the citadel would endure even when multiple hits had riddled the ends. The Victoria collision was to show that doors, ventilation and valves do not remain tight after damage and Inflexible would probably have foundered from slow flooding into this citadel. Barnaby claimed that she was designed to withstand a torpedo hit with the centreline bulkhead giving only a small heel – but he did not envisage flooding extending beyond one transverse compartment.

However, it is difficult to see a better solution to the design requirement and the concept received some vindication from the battle of the Yalu Sea on 17 September 1894 when two Chinese ironclads, Ting Yuen and Chen Yuan, to Inflexible’s configuration, but smaller, received a very large number of hits and survived. To some extent, the 1913 trial firings against the Edinburgh may be seen as justifying the concept. Opponents of the Inflexible mainly favoured protected cruisers whose only protection was similar to that at the ends of the Inflexible which they derided. White gives her cost as £812,000 though other, much lower, figures have been quoted. There were two diminutives which call for no mention.

‘The Ship is a Steam Being’

Reed’s letter, quoted at the beginning of the chapter, referred to the increasing use of auxiliary machinery. Some early examples include; a capstan in Hercules (1866), hydraulic steering gear, fitted to Warrior in 1870, and a steam steering engine for Northumberland as well as the turrets in Thunderer and later ships. The number increased rapidly and Inflexible was truly a ‘steam being’. Her auxiliaries comprised:

1 steering engine

2 reversing engines

2 vertical direct fire engines

2 pairs steam/hydraulic engines to work the 750-ton turrets

1 capstan engine

4 ash hoists

1 vertical direct turning engine

2 40hp pumping engines, total capacity 4800 tons/hr 2 donkey engines for bilge pumping

2 steam shot hoists

4 auxiliary feed, similar to donkey engines.

2 Brotherhood 3 cylinder for boat hoisting

4 Brotherhood 3-cylinder fan engines

4 Friedman ejectors

2 horizontal direct acting centrifugal circulating pump

The list above does not mention ventilation fans but it is virtually certain that these were fitted. It was some time before satisfactory ventilation systems were developed. An electric searchlight was tried in Comet in 1874 and the first permanent fitting was in Minotaur in 1876. Inflexible had 800-volt d.c. generators by the US Brush company. These powered arc lights in the machinery space and Swan ‘Glow’ lamps elsewhere. The Swan lamps were connected in series and it was a year before the 800-volt system killed its first victim. She was even launched by electricity; when Princess Louise touched a button, a wire fused and the bottle of wine fell and weights crashed onto the dog shores.

Fire Direction and Radar Equipment on the Bismarck Class BBs

Weapons and Fire Control Systems

The designers of the Bismarck class adhered to the tried and tested main armament arrangement of two twin turrets forward and aft, the rearmost of each superfiring. The reason for this was the better field of fire and more effective sequence of salvos. The smaller calibres—the 15cm secondary artillery and the 10.5cm flak—followed the previous layout.

The concept of the 15cm gun was its role as a classic anti-destroyer weapon. It fired a theoretical eight, but in practice only six, rounds per barrel per minute, and was in no respects of any value as an anti-aircraft gun, having too slow a rate of fire and turret rotation speed and an inadequate angle of elevation. Together with the main armament, it was used on Tirpitz in an anti-aircraft role as it could put up a long-range barrage of time-fused shells to confront approaching bomber formations with a curtain of shrapnel.

German naval flak was inadequate, and lacked a gun which was capable of engaging both a fast bomber at high altitude and long distance and also a torpedo bomber closing in just above the wave tops. The planners had failed to grasp the concept of the multi-purpose flak gun. There would certainly have been room for them, but it was left to other navies to address the problem and to come up with workable solutions towards the end of the war. Of course, Germany already had an excellent flak gun, the 12.7cm Flak L/45 Model 34, which had a range, at 30 degrees’ elevation, of 10,497.3m, a shell weight of 23.45kg and a muzzle velocity of 829.97m/sec and which had given outstanding results against enemy bombers over the Reich.

The VDI-Memorandum (which had had handwritten comments added in April 1957 by former ministerial adviser Dipl-Ing Ludwig Cordes, from December 1942 Chief of the Official Group for Artillery Construction at Naval Command, a personality familiar with the whole subject inside and out) drew special attention to fire direction centres with the following notable conclusion:

There was no technical expert at Naval Command (OKM) charged with responsibility for this particular interest. Rulings were ultimately within the jurisdiction of a military centre, which led to frequent erroneous decisions.’


There were two different models. The 10.5cm model C33 guns of Bismarck were fitted in twin mountings, C31 forward and C37 aft. The guns differed principally in the coordination system for their target data. In themselves both weapons were flawless, but unfortunately when the C37 had been shipped, the necessity to install the fire direction equipment individual to each model of gun had been overlooked, with the result that, when the fire direction instructions were transmitted, Flak C33 fired at the target and Flak C37 at a point beyond it. The error here clearly lay with Kriegsmarine planning, which resulted in the linking of an incompatible battery to the control centre.

Flak Direction Centres

Until the end of the war, German heavy units were equipped with grossly inferior flak direction centres based on the Cardan ring system with a large revolving base. At a massive 40 tons, their weight tended to affect the ship’s stability. In battle, many defects came to light, for the Cardan ring system was very sensitive to underwater hits: even the lightest hits could cause a break in the ring, resulting in a total system breakdown.

As early as 1932 engineers had set out proposals for an improved and more suitable development which had a smaller and triaxial rotating base. Despite repeated reminders, it was not until 1942 that the new device was first commissioned, and the experimental prototype was eventually ready by the end of the war though never fitted aboard ship. Complementing a far superior handling capability and better armour protection, the new device had a weight of only 6 tons.

In 1933 proposals had been put forward for automatic fire direction mountings for 3.7cm and 2cm guns. This demonstrates how far-sighted the German weapons engineering industry was, but in this case nothing came of the proposals.

Radar Equipment

At the outbreak of war in 1939 Germany had two workable radar systems, Freya (2.4cm waveband) and Würzburg (50cm waveband). At the time, the Third Reich led the world in this field. This would change. In the autumn of that year the British built a 12m system and then concentrated their efforts on the centimetre wavebands. In 1943, they introduced the 9cm device known to the Germans as ‘Rotterdam’.

In Germany the industry was fragmented, and instead of drawing on the experience of well-established firms, new companies were set up and the Luftwaffe commandeered all new developments. In 1942–43 it was decided that no new developments in radars of wavebands less than 20cm were possible, and all research into that area was abandoned. Only when a ‘Rotterdam’ set fell into German hands was work resumed. None of the equipment built worked satisfactorily in service. Germany had ‘missed the bus’.

These few concluding remarks may be sufficient to permit a more critical assessment of German warship construction of the period than is normally the case. At their completion, the two Bismarck class units were the culmination of capital ship building, but they were already obsolescent. They were powerful and sturdy fighting ships, but not unsinkable. In their final form they were, asthetically, the crowning glory of German warship construction.

The destruction by Bismarck of the world’s largest capital ship of the time, the battlecruiser Hood, is an impressive testimonial to German naval gunnery. But in respect of this success, it must be remembered that it was achieved against a warship which had been laid down in the Great War twenty-five years previously—certainly modernised but unchanged in her basic structure.




Preliminary reconstruction of one of Khubilai Khan’s lost ships. The result of generations of Chinese engineering and development, these were the world’s most advanced warships duringthe Medieval period. He squandered his naval advantage with poorly executed attacks on Japan, Vietnam, and Java.

Khubilai Khan’s Lost Fleet

In Search of a Legendary Armada

by James P. Delgado (Author)

On October 19, 1274, a massive Mongol war fleet sailed into Hakata, Japan’s most important harbor for overseas trade. Chinese records of the time claim a thousand ships and more than twenty-three thousand soldiers, though modern scholars believe that the actual numbers of both ships and soldiers were considerably smaller. To the beat of huge war drums the Mongols and their allied Korean troops came ashore in small landing craft. News of the imminent invasion had well preceded the fleet’s actual arrival, and a substantial force of samurai, at least six thousand, awaited them.

Hand-to-hand combat began on the beach. Both sides took heavy casualties. Japanese sources claim that two thousand samurai died on the beach and in the pine grove adjoining the shore. The Mongol forces gradually pushed the samurai back into Hakata town. Fighting continued in the streets and alleys. By nightfall, the invading troops had taken and burned the port. The defending samurai regrouped in the hills above the town.

Through the early hours of night the commanders of the Mongol/Korean force debated tactics. One faction favored an immediate night attack to press their advantage. Other commanders argued that the troops were exhausted and needed sleep. Finally, it was decided to continue the battle in the morning, and the troops returned to their ships. In the morning, however, the fleet was gone from Hakata Bay. Japanese sources report that a strong, “divine” wind blew the ships out of the harbor and into the sea.

The likeliest scenario is that the fleet simply sailed away, its commanders aware of problems that the Japanese were not. The fleet was low on arrows, having used large numbers in taking two strategic islands on the way to Hakata. The commanders perhaps also wanted to reconsider their strategy. Struggling ashore and fighting hand to hand on a beach and in trees was probably the least favorable terrain for Mongol troops. They were superb cavalry, trained for plains battles, massed arrow attacks, and group maneuvers, but largely untrained in hand-to-hand sword fighting on foot, and avoided this sort of battle whenever possible.

The results of the first battle of Hakata were perhaps satisfactory to the great Mongol ruler Kublai Khan, Genghis Khan’s grandson. His strategy was straightforward: conquer all China and supplant the Song dynasty. By and large, the war was going well. Mongol armies had pushed the Song into far southern China. The destruction of Hakata meant that the Song would gain no revenue from trade with Japan.

This first battle of Hakata, however, produced no shipwrecks. Even the Japanese sources concede that only a few of the Mongol ships were beached by the mysterious wind that blew the fleet back to “their lands.”

Much had changed between the first invasion attempt in 1274 and the second invasion in 1281. Mongol armies had pursued the remaining Song forces into South China, defeated them, and captured and executed the last emperor. Kublai Khan was, indeed, ruler of a united China, with all the resources and the problems that entailed. He founded a new dynasty, the Yuan, and moved his capital from Karakorum, deep in Mongolia, to Beijing, the better to rule his new conquests.

Kublai Khan sent envoys to Japan, in 1279, demanding surrender. The bakufu, head of the alliance of Japanese nobles, had the envoys executed on the beach at Hakata. Kublai Khan and the king of Korea conferred and agreed the invasion force to conquer Japan would consist of one hundred thousand troops. The king of Korea agreed to construct an enormous fleet, which would carry Mongol and Korean troops across the Korea Strait to Hakata. Kublai Khan ordered a second fleet constructed on the Chinese coast, which would carry Chinese troops to join the Koreans and Mongols at Iki Island off Japan’s west coast.

For more than a year, in both Korea and south China forests were stripped for the ships and harsh taxes levied to equip them. The Koreans, eager to engage, sailed in early May 1281, knowing that the Chinese fleet was not ready. The samurai had constructed a stone wall along the beach at Hakata, which halted the invading force. In heavy fighting the samurai drove the Mongols and Koreans back to their boats. A stalemate set in, the samurai holding the beach and the port and the Mongols and Koreans holding the harbor. The samurai attacked the fleet in small boats, sometimes boarding, sometimes pushing fire-rafts to burn the invader’s ships. The attacks eventually forced the invading fleet into a compact defensive circle in the bay.

The Chinese fleet eventually did arrive but could not assist in the stalemate at Hakata. Instead, the Chinese attacked inland from Imari Bay, thirty miles south of Hakata. Samurai fought the Chinese soldiers in the inland hills, finally pushing them back to their ships. In the end a typhoon destroyed both fleets, which were at anchor through the height of the typhoon season. The fierce storm piled ship upon ship, driving them onto the rocky shore. Casualty estimates are, of course, speculative but run upward of fifty thousand men. Some thirty thousand Chinese soldiers were captured and enslaved. Both Chinese and Japanese sources agree that the second battle of Hakata Bay littered the bottom with wreckage.

The Mongols at War

The two opponents at Hakata Bay had quite different military and political backgrounds. Fifty years earlier Genghis Khan had reorganized bands of steppe cavalry into the most successful rapid strike force the world had ever seen. The important changes were in organization, discipline, and ideology. Genghis Khan reassigned the men of family and ethnic units into mixed units, thereby promoting loyalty to the larger Mongol goals rather than narrow family concerns. The units were arranged on a decimal system, with commanders over one hundred, a thousand, and ten thousand men. Cavalry practiced daily and honed their skills in frequent large hunts. Genghis Khan also enforced discipline on the welter of ethnicities that constituted his army. For example, looting after battle was prohibited on pain of death. The military goal was to annihilate the opposing force, and looting disrupted the process. Genghis Khan promulgated and practiced his belief in “world conquest”—his forces were destined to defeat all opposition and rule the entire world. This ideology is perhaps best exemplified by a letter from Guyuk, grandson of Genghis Khan, to Pope Urban IV. The pope, in an official letter, proposed an alliance between the European kings and the Mongols against Muslims, as their common foe. Guyuk replied:

Thanks to the power of the Eternal Heaven, all lands have been given to us from sunrise to sunset. How could anyone act other than in accordance with the commands of Heaven? Now your own upright heart must tell you: “We will become subject to you, and will place our powers at your disposal.” You in person, at the head of the monarchs, all of you, without exception, must come to tender us service and pay us homage; then only will we recognize your submission. But if you do not obey the commands of Heaven, and run counter to our orders, we shall know that you are our foe.

Mongol forces were mounted cavalry and used a short reverse-curve bow, which could be shot from horseback. With both hands occupied with the bow and arrow, Mongol cavalry had to control their horses with their knees, commands every horse knew and every horseman practiced from childhood onward. The reverse-curve bow was of composite materials, including wood, horn, and steel. It was enormously powerful, capable of penetrating armor at 150 yards. The preferred tactics of Mongol cavalry therefore avoided charges into well-entrenched positions. They much preferred tactics that included massed arrow attacks from outside the range of enemy weapons; the feigned retreat, which drew the enemy into ambush; or large-scale flanking movements, which resulted in attacking the enemy on three sides. These maneuvers depended on careful tactical coordination, usually by means of large signal flags. Mongol armies were, therefore, at their best in plains battles, with room to maneuver their horses and sweep in large formations.

Commanders of opposing forces quickly learned that they would likely lose a plains battle to Genghis Khan. Those who could, retreated to fortified positions. Genghis Khan’s first siege was in 1218 at Otrar, a typical Silk Road fortified town in what is now southern Kazakhstan. After establishing friendly relations with the king of the region, Genghis Khan equipped and financed a large caravan of Muslim traders to buy luxuries on the Silk Road and bring them for sale to his capital. Four hundred and fifty Muslim traders purchased silks, satins, carpets, and gems. When the returning caravan halted at Otrar, the governor of Otrar seized the goods and animals and executed the traders. In the colorful language of the Secret History of the Mongols (written shortly after Genghis Khan’s death),

The control of repose and tranquility was removed, and the whirlwind of anger cast dust into the eyes of patience and clemency while the fire of wrath flared up with such a flame that it drove the water from his eyes and could be quenched only by the shedding of blood. In this fever Cheingiz-Khan went alone to the summit of a hill, bared his head, turned his face toward the south and for three days and nights offered up prayer, saying: “I was not author of this trouble; grant me strength to extract vengeance.”

Genghis Khan divided his army, half attacking in the north of the kingdom to tie down the king’s forces, the other half investing Otrar, which had been reinforced with thousands of royal troops. Genghis Khan had no clever siege engines, no catapults or trebuchets, only tenacity. The army formed “several circles around the citadel,” fought the sallies from the city, and maintained the siege for five months. In desperation some of the town’s troops rode out and offered service to Genghis Khan. He saw their action as dishonorable and executed them as his troops poured through the undefended gate. “All the guilty and innocent of Otrar, both the wearers of the veil and those that donned kulah and turban, were driven forth from the town like a flock of sheep, and the Mongols looted whatever goods and wares were there to be found.” The Mongol troops eventually fought their way into the citadel and captured the offending governor alive. He was executed by pouring molten silver down his throat, just punishment for his greed.

Though the Mongols are famous for their sweeping cavalry strategies, a majority of Genghis Khan’s battles were actually fought against a fortified hill, palisade, or town. The Mongols quickly copied from their opponents a weapon of war new to them, the trebuchet, which utilized a heavy counterweight’s force multiplied by a long lever arm and an equally long flexible sling. Invented either in Europe or the Muslim West (though perhaps an improvement of an earlier Chinese catapult), the trebuchet hurled a heavy stone (generally more than 150 pounds) with enormous force, capable of knocking down men and horses like bowling pins and equally capable of crashing through gates and walls. Genghis Khan recruited and gave military appointments to Muslim technicians capable of building such a weapon.

Less than two decades later Mongol siege engines from the West and the technicians to build them had moved across all Asia and were attacking fortified cities in China. Only three years after Otrar, the Mongols were using siege engines on the eastern front in their campaign against the fortified cities of northern China. Thus, it is no surprise that the Mongols took great, fortified cities. Baghdad, one of the largest cities in Asia at the time, fell to the Mongols in 1258 (fifteen years before Kublai Khan attacked Japan). It is likely that the great Mongol fleet that attacked Hakata Bay carried siege engines such as the trebuchet in anticipation of attacking forts and fortified cities.

Mongol armies generally suffered defeats in only two circumstances. First, highly trained professional soldiers who knew Mongol strategy and tactics occasionally simply outperformed them. The Mamluks, full-time, trained slave-soldiers, were just such a force and defeated the Mongols in Egypt. Second, problems of adverse terrain limited the effectiveness of Mongol cavalry. Mountains were a serious problem for the Mongols. Horsemen could not wheel and move in large units. Ambush lurked in every defile. Even in defeat the enemy could disappear into the mountains, eliminating the Mongol tactic of annihilating the opposing army. Massed arrow attacks did little against mountain fortresses, which were also almost impossible to surround. Troops from the fortresses could often defend agricultural land nearby, which provided the fortress with food. The combination of mountains, fortresses, and resolute resistance, for example, made the conquest of Sichuan, a southwestern province of China, slow, difficult, and costly. Mongols fought in the mountains of Sichuan virtually every year for more than three decades before conquering it.

China’s coastal plain was equally difficult terrain for Mongol armies. Canals crisscrossed it, and the rice fields were flooded much of the year. Large-scale cavalry movements were impossible. Fortified cities were frequent and were connected by boat more than road. The Mongols had to adapt, and they did, incorporating Chinese and Korean leaders and infantry who knew how to fight in this watery terrain, so different from the dry steppe of the Mongol homelands. Mongol armies traveled by boat and learned siege techniques. They recruited artisans to build the powerful Chinese trebuchet. Chinese troops used gunpowder weapons extensively for the first time.

Samurai Warriors

On the beach at Hakata Bay were six thousand of the most highly trained, most professional, and best-equipped troops the Mongols ever faced. Samurai were the elite product of an entire social and economic system, just as were the Mongols. Within the fragmented Japanese political system, wars between elite families were frequent, and formal training in schools of the martial arts was mandatory for elite men (and a few elite women). A nineteenth-century text of one of these schools well illustrates the focus and rigor of samurai training. Students learned, for example, unarmed fighting, grappling, short sword fighting, quick sword drawing, stick fighting, dagger technique, the use of rope, and crossing rivers in armor on horseback. The training was as much mental as physical:

Because the beginner does not know how to stand with the sword in his hands or anything else, in his mind there is not a thing to be attached to. When he is attacked, without any deliberation he tries to fend off the attack. But gradually he is taught many things, he is instructed how to hold the sword, where to concentrate his mind and other things. So his mind will be attached to those things and when he attempts to attack his opponent, his movements will be awkward. However, as days, months and years pass, due to innumerable trainings, everything, as he stands, as he holds the sword will lose consciousness, in the end getting back to the state of mind he had in the beginning, when he did not know anything.

The samurai code of honor preferred single combat, which was almost certainly a detriment in their first encounter with the Mongols. Samurai quickly learned that Mongols were quite content to fire massed arrows at any opponent who sought single combat. The samurai also learned that their superior sword skills made up for lesser numbers in close combat. A recent scholarly book has persuasively argued that the samurai needed no “divine wind” to drive off the Mongol ships. They repelled the invasion based on their skills, armor, and training.

Shipbuilding in the China Sea

What sort of ships brought the Mongol invasion fleet from Korea to Japan? The evidence is meager but suggests that Korean long-distance trade ships were the likeliest carriers. The decorative back of a lady’s mirror from the period shows such a Korean ship, sails reefed, in roiling seas. Recovered timbers and planks of actual vessels show that these craft had an almost flat bottom. Shipbuilders attached successive planks of pine with overlapping edges and mortise-and-tenon joints. Elm was used for pegs to lock the mortise and tenons in place. Oak was used for a heavy yoke, which was set amidships and served as a sturdy cross member to stabilize the hull. Cross planks of oak were fitted low in the hull for the same purpose. Another layer of heavy oak crossbeams joined the upper planks of the two sides of the hull. The pattern of crossbeam support passing through the planks was apparently unique to Korea. Xu Jing, a Chinese emissary to the court of Korea, noted that the Korean ships were different from contemporary Chinese craft.

Both Chinese and Korean long-distance ships had a stern rudder, a large mast set amidships, and a smaller foresail. Sails were rectangular and reinforced with battens. Chinese and Korean ships used a windlass to raise the heavy anchor (as the scene on the Korean mirror shows). Korean ships had a planked deck, but it is unknown whether the space below the deck was divided into holds, as was typical of Chinese ships of the period. The mirror scene shows piled goods on deck and commodious cabins for the rich merchants who owned the goods. Korean sources assert that seventy people could comfortably sail on these ships. The current state of the archaeological, textual, and visual evidence does not permit even a speculation on the size and tonnage of these craft.

About the Chinese ships, which formed the second fleet attacking Japan, we have good material evidence. In 1974, Chinese archaeologists excavated a hull from the mud off Quanzhou Bay. The ship was amazingly intact from the waterline down. Coinage aboard dated the ship to 1272, only two years before Kublai Khan’s first attack on Hakata Bay. The ship was 113 feet long, with a beam of 32 feet, drew only 10 feet of water, and displaced about 375 tons. Unlike stereotypical Chinese ships with flat bottoms and ends, the Quanzhou ship had a keel, was V-shaped in section, and had sharp prow. Twelve bulkheads divided the hull, which also had stepping for three masts. A flat transom carried the rudder, rather than a sternpost. Iron nails secured the overlapping planking. The cargo of incense wood, pepper, and hematite suggests that this was a long-distance goods carrier, returning from Southeast Asia. Such a ship could have been impressed to carry troops to Japan.

In the last three decades Japanese archaeologists have been searching Hakata Bay for the physical remains of the battle of 1281. Tantalizing evidence has turned up, such as Chinese- and Korean-style anchors, Chinese ceramics, disc-shaped articulated armor, and weapons typical of Mongol fighters. Various scans of the bottom of the bay have revealed clumps of timbers, which are likely the remains of a ship or the mixed remains of several ships. Much of the timber is smaller than that used in big Korean trade ships, which suggests that the Mongols also commandeered coastal craft and probably even flat-bottomed river craft.

Archaeologists in 2013 located a section of an intact hull. Ultrasound scans revealed a thirty-six-foot section of keel with adjoining planking under only three feet of sediment just off the shore in Hakata harbor. Ceramics, stone anchors, and other artifacts surround the wreck. For now, it remains buried, awaiting future excavation.

In a larger geopolitical perspective, Japan, Korea, and the east coast of China formed a complex a maritime world, which was roughly the same size as Europe’s northern littoral. From Nagasaki, Japan, to Shanghai, China, across the Yellow Sea is five hundred miles, about the same distance as Scandinavia to England. Korea and Japan are only one hundred miles apart, roughly comparable to the twenty-five miles that separate England and France across the Channel. Over the centuries, just as the Scandinavians invaded England and the English used their ships to invade the French, so too did Chinese, Korean, and Japanese dynasties invade each other’s territory, trade with each other, sponsor piracy of each other’s shipping, ally in attacks on each other, call in each other to put down indigenous rebels, and constitute places of refuge for defeated or aspiring rulers.

Dynasties of Korea, Japan, and China sometimes chose to close their maritime borders, forbidding traders from entering and citizens from leaving. These legal prohibitions typically were not effective. Traders and travelers found ways to circumvent them. As also happened in Europe, local or regional powers in the China Sea region founded new ports beyond the reach of the central government. One of the most famous of such ports was Hainan Island off the southern coast of China, which served smugglers at the time of the Kublai Khan expedition and for several subsequent centuries.

Since the history of China is usually written as the history of dynasties, we might assume that the royal court of China was always the dominant power on land and at sea, but this is simply not the case. Periods of warring states were as frequent as periods of stable, large dynasties. The south of China was always difficult for a northern-based dynasty to integrate. Declining dynasties sometimes looked across the seas for a Japanese or Korean alliance.


Bangor-Class Minesweeper (reciprocating engine): A wide minesweeper class-it formed a large part of the Neptune minesweeper strength of 287. Bangor-class ‘sweepers were built in three versions-diesel, turbine, and reciprocating engines. Displacement: 672 tons. Crew: 60. Speed: 16 knots. Armament: 1 x 3-in, 1 x 40-mm, 4 x .303-in mg

RAF Air/Sea Rescue Launch: These boats saved about 13,000 airmen during the war, and were the natural seaborne counterpart to the huge Allied air umbrella which would cover the Neptune operations. Length: 68 feet. Speed: 38 knots. Armament: 3 x .303 Lewis mg

Armed Salvage Tug: Quite apart from the job of moving the huge Mulberry units to the Normandy coast, the fleets of tug-boats had to cope with about half the landing-craft -those unable to cross the Channel under their own power, and which could not be carried on the decks of transports. Displacement: 700 tons. Speed: 13 knots. Crew: 30. Armament: 1 x 3-in, 2 x 20-mm, 2 x .303-in mg

Armed Trawler: Trawlers-with the minesweepers and other light flotilla craft-were to perform invaluable services as convoy escorts shepherding and marshalling the transports. Hundreds of them were pressed into service for this role. Typical specifications were: Displacement: 378 tons. Speed: 11 1/2 knots. Crew: 30. Armament: 1 x 4-in, 3 x 20-mm, 2 x .303 mg; about 20 depth-charges


The ‘Mulberry Harbours’ was a WW2 civil engineering project of immense size and complexity. The floating harbours provided port facilities during the invasion of Normandy from June 1944 until French ports like Cherbourg were captured.

To carry the overall total of 40-50,000 men with their vehicles and equipment, an armada of over 4,000 landing ships, landing-craft, and barges of varying types was required; only about half of these were capable of crossing the Channel under their own power, the remainder having either to be towed or carried aboard the larger ships. When it is remembered that every man and every vehicle had -to be allotted to a specific ship or landing-craft, that every vehicle before embarkation had to be waterproofed, that men and vehicles had to arrive at the right time at the ‘hards’ (improvised landing places) at which their particular landing-craft was beached – the complications of planning and organisation can easily be imagined. Only when all this had been done, did the huge task of the Royal Navy in assembling, marshalling, and shepherding this heterogeneous collection of vessels across the Channel into paths swept through the enemy minefields, in landing them on the correct beaches, and providing the necessary fire support begin.

Apart from the non-combatant ships and landing-craft carrying men, vehicles, and stores, the US Navy and Royal Navy assembled for the escort and support of the operation a fleet of over 1,500 vessels, ranging from battleships to armed landing-craft. They were divided into two ‘task forces’, the Western Task Force from the US Navy supporting the US landing, and the Eastern Task Force, provided primarily by the Royal Navy, supporting the British/ Canadian landing. Each task force was further sub-divided into ‘forces’, one to each beach, responsible for escorting the assaulting force concerned, positioning it correctly, and providing fire support for the landing. The assault looked primarily to naval gunfire to silence the German coastal batteries and strongpoints.

The magnitude of the naval effort can be judged by the fact that the forces included seven battleships, 23 cruisers, 148 destroyers, as well as a swarm of smaller vessels -sloops, frigates, trawlers, corvettes, patrol craft and minesweepers. In addition, a fleet of 350 specially designed landing-craft carrying guns, rockets, anti-aircraft guns, and machine-guns was assembled for the close support of the actual assault.

The air effort in direct support of the assault was on a similar immense scale. It comprised the most modern types of aircraft available at the period, primarily the Spitfire, Mustang, Typhoon, Lightning, and Thunderbolt.

The exact timing of the assault proved a most complex problem which had its repercussions upon the dates on which the operation could be launched.

Although from the point of view of the assaulting troops there was much to be said for an assault in darkness, both the navies and air forces had to have daylight to carry out their bombardment tasks, and darkness would dangerously increase the likelihood of troops being landed in the wrong place. To assist navigation and for the airborne landings, moonlight was essential. Finally, the German underwater beach obstacles meant that landing must begin three to four hours before high tide. The only suitable periods for the operation therefore were those when there was four to five hours’ daylight between dawn and high tide and at the same time good moonlight was available. All these conditions could only be satisfied on approximately three days in each lunar month.

However powerful and successful the assault, it would clearly be valueless unless the forces ashore could be built up more rapidly than those of the enemy and properly maintained when there. This involved three problems: the planning, escorting, and routing of the follow-up convoys; ensuring that those convoys contained the right personnel, vehicles, and equipment arriving in the right order; and finally ensuring that they could be rapidly unloaded on arrival.

Fifteen personnel ships, 74 ocean-going merchant ships, and more than 200 coasters were loaded before D-Day and these were to form the first wave of the build-up; the requirement thereafter was for eight convoys a day. Once having got the assault force across, however, movement of these convoys was not likely to present any particular problem for the Allied navies.

The question of what they should contain after the initial preplanned flight was more difficult, and a special organisation known as ‘Build-Up Control'(BUCO) was set up in Southampton to ensure that what was shipped across the Channel was geared to the requirements of the battle.

The question of rapid unloading initially appeared the most difficult of all; it could clearly not be done across the beaches as a long-term measure and the likelihood of capturing port facilities intact appeared small, at any rate in the early stages. The problem was solved by perhaps the most famous devices of the entire operation-the artificial harbours known as ‘Mulberries’. They owed their existence primarily to the foresight of Churchill himself, who had directed their development as early as 1942, with his oft-quoted minute: ‘They must float up and down with the tide. . . . Don’t argue the matter. The difficulties will argue for themselves.’ They consisted of an outer breakwater formed partly of sunken blockships and partly of concrete ‘caissons’, 200 feet long, which had to be towed across the Channel; in the area of sheltered water so created were floating piers adapted to take coasters, landing ships or barges; unloading was further assisted by a fleet of amphibious lorries known as DUKWs. The success of the system may be judged by the fact that shortly after the assault, an average of 6,500 vehicles and nearly 40,000 tons of stores was being landed weekly.

The supply of motor and aircraft fuel presented a particular problem. Initially tankers were moored offshore and the fuel fed by buoyed pipeline into depots on land. Preparations were made, however, for an underwater pipeline direct from England to the French coast-PLUTO, or ‘Pipe-Line-Under-the-Ocean’-and eventually, though not in the early stages, fuel supply was in effect drawn direct from England.

The impression so far given may well be that this was an exclusively British/American/Canadian operation, but the contributions of the other Allies must not be forgotten. The French, for instance, provided two cruisers, one destroyer, one armoured division, and four squadrons of aircraft; the Belgians one brigade and two squadrons; the Dutch two gunboats, one brigade, and two squadrons; the Poles one cruiser, two destroyers, one armoured division, and nine squadrons; the Norwegians three destroyers and four squadrons; the Czechs three squadrons; the Australians five squadrons; and the New Zealanders five squadrons. Practically every occupied country of Europe was represented in one way or another.

Finally, it must not be forgotten that the invading forces could expect assistance from an ally inside France. In this context the French resisters deserve to be included in the catalogue of forces available to the Allies. SOE had been its best to organise and arm them and from early 1943 German demands for labour had assisted recruiting; by 1944 some 100,000 young men had taken to the ‘Maquis’. The vast majority of the resisters’ plans were, of course, geared to the great day when the Allies would land in France once more; in 1943 they had put forward a series of seven ambitious plans to deal with railways, road movement, telecommunications, ammunition dumps, oil fuel installations, headquarters, and railway turntables. By the beginning of 1944, however, many resistance networks had been broken up by the Gestapo and only the railway demolition plan appeared to be capable of any certain implementation. Nevertheless both numbers available and arms supplied were considerable: by May 1944 80,000 Sten-guns, 30,000 pistols, 17,000 rifles, and nearly 3,500 Bren-guns had been parachuted into France; overall there were probably some 100,000 men plus another 35- 40,000 in the Maquis who had a weapon of some sort.

To assist the Resistance and ensure that as far as possible its operations were co-ordinated with those of the Allies, SOE prepared a number of three-man teams (American/British/French) to be parachuted-in uniform -into areas where resistance was expected to flourish, so as to act as liaison between the resisters and the regular forces. In addition the British Special Air Service and the American Operational Groups were entrusted with various raiding and harassing operations for which it was hoped that they could obtain the assistance and support of the Resistance.

Finally, preparations had to be made to take over and run civil affairs in the liberated areas of France pending recognition of a French government. Here also assistance from the Resistance organisation was hoped for.

So the balance sheet of this immense operation shows the following staggering figures:

  • 50,000 men in the assault drawn from five divisions;
  • Over 2,000,000 men to be shipped to France overall, comprising’ a total of 39 divisions;
  • 138 major warships used in the assault, together with 221 smaller combat vessels (destroyer category and below);
  • Over 1,000 minesweepers and auxiliary vessels;
  • 4,000 landing ships or craft;
  • 805 merchant ships;
  • 59 blockships;
  • 300 miscellaneous small craft;
  • 11,000 aircraft, including fighters, bombers, transports, and gliders;
  • Over 100,000 partially armed men of the Resistance ready to lend such support as they could.

With such a weight of numbers and material it might well be thought that the assault would be practically irresistible, but two factors must be remembered: first, the hazards and extreme complexity of an amphibious operation of this magnitude, for which there was no precedent; all the most meticulous arrangements could be upset and the utmost confusion caused by some chance occurrence or unpredicted change in the weather: second, the great inherent superiority of the defence over the attack in an amphibious operation, especially against prepared coastal defences; however great the Allied superiority, there could be no certainty that the force would succeed even in securing a foothold. Surprise and deception were the essence of the operation-and complete surprise there could not be, for the Germans knew that the invasion was coming; all they did not know was when and where.

Finally, it was clear to both sides that this was the decisive operation of the war. If the landing succeeded, Germany must eventually be crushed sooner or later between the advancing forces of the Russians and the Allies. Should the landing fail, the Allies might well take years to recover from their losses in men, material, and morale; the peoples of Occupied Europe would give up hope and Germany would be left free to turn and square the account with the Russians.

The broad shoulders of General Eisenhower carried an immense burden.

‘X Lighters’

A number [100] of large purpose-built `X Lighters’ had been developed by the Navy in the First World War and these contributed to the landings at Suvla Bay in 1915.

Although something of a military sideshow, the campaign at Gallipoli also made a contribution to the acceptance of the internal combustion engine for marine use. Anticipating the need to land troops and equipment on beaches here and elsewhere, the British Admiralty ordered large numbers of small landing craft. To ensure a shallow draught, these X-lighters were fitted with lightweight motors, many being hot-bulb engines. After the war, these X-lighters were sold off and many were converted for commercial use, giving some British coastal shipowners their first experience of the motor vessel.

The search for a more suitable landing craft continued after the war and this requirement was consistently emphasised in exercise reports. In the 1920s an inter-service Landing Craft Committee was established to study the design and number of craft required to conduct a landing on a hostile shore. Their first attempt at a landing craft was the Motor Landing Craft (MLC(1)) completed in 1926. This craft was not a success and was followed in 1928 by the MLC(10). The MLC(10) was a flat-bottomed craft powered by a water jet. It could embark 100 troops or a 12-ton tank, discharging them directly onto the beach via a steep bow ramp. The water jet gave it a relatively slow speed of only 5 knots and the boat’s flat bottom and bow ramp made it rather unseaworthy, handicaps that are common in modern amphibious craft. By 1934 the MLC had been thoroughly tested in a series of exercises and the design proved satisfactory. Two more vessels were procured and these were joined by six more, ordered as a result of the 1936 Abyssinian crisis.

X-Lighters in WWI and at Gallipoli

X Lighters

Displacement 200 tons
 Dimensions  105ft 6″ x 21 x 7ft 6″
 Guns Unarmed (The crew may have had side arms for self-defence or covering fire on the beaches)
 Machinery  steam or diesel engines, speed 8 kts
 Crew  4
 Builders  various
 Laid Down  1915
 Completed  1915- 18

American Civil War Ironclads

At the outset both sides were militarily weak. The North did have a clear advantage at sea, although its widely scattered force of 80 warships was totally inadequate for what lay ahead. On 19 April Lincoln proclaimed a blockade of the 3,500 miles of Confederate coastline. Secretary of the Navy Gideon Welles launched a major construction program, which included ironclads. Washington also purchased civilian ships of all types, many of them steamers, for blockade duty.

In April 1861, upon the secession of Virginia, the South gained control of the largest prewar U. S. Navy yard at Gosport (Norfolk) along with 1,200 heavy guns, valuable naval stores, and some vessels. Among the latter was the powerful modern steam frigate Merrimack. Set on fire by retreating Union forces, she burned only to the waterline before sinking. The Confederates raised her and rebuilt her as the ironclad Virginia. Confederate Secretary of the Navy Stephen Mallory hoped to offset the Northern naval advantage by ironclad warships capable of breaking the blockade, and he advocated commerce raiding, the traditional course of action of a weaker naval power against a nation with a vulnerable merchant marine. Mallory hoped to drive up insurance costs, weaken Northern resolve, and force the U. S. Navy to shift warships from blockade duties

Each side also constructed ironclads. The first were actually built by the Union to help secure control of America’s great interior rivers. Thanks to its superior manufacturing resources, the Union got its river fleet built quickly. In August 1861 the army ordered seven ironclad gunboats. Constructed by James B. Eads, they were the first purpose-built ironclad warships in the Western Hemisphere.

The so-called Peninsula Campaign set up history’s first battle between ironclads. On 8 March 1862 the Confederate ironclad Virginia sortied from Norfolk and sank two Union warships. That evening the Union ironclad Monitor arrived, and the next day the two fought an inconclusive battle, which nonetheless left Union forces in control of Hampton Roads. “Monitor fever” now swept the North, which built more than 50 warships of this type. The Confederates countered with casemated vessels along the lines of the Virginia, the best known of these being the Arkansas, Manassas, Atlanta, Nashville, and Tennessee. Also, the Confederacy secretly contracted in Britain for two powerful seagoing ironclad ships. These so-called Laird Rams were turreted vessels superior to any U. S. Navy warship, but when the war shifted decisively in favor of the Union the British government took them over.

Union Monitors

The distinction for participating in the first ironclad-to-ironclad clash must go to the Ericsson turret armorclad USS Monitor, the world’s first mastless ironclad. At the Battle of Hampton Roads (8 March 1862), Monitor faced off Confederate ironclad battery CSS Virginia in one of the very few naval battles fought before a large audience, lining the Virginia shore.

It is popularly supposed that Hampton Roads demonstrated that the day of the wooden warship had ended. It did no such thing; the armored Kinburn batteries had already taken the world’s attention almost six years before, the French La Gloire had been in service for the previous two years, and the magnificent seagoing British ironclad HMS Warrior for six months; and the world’s naval powers at the time had some 20 ironclads on the stocks. It would have been a peculiarly dense naval officer or designer who did not realize by March 1862 that ironclads would dominate the world’s fleets in the very near future. The main question would be what forms those ironclad warships would take.

The historic Battle of Hampton Roads did touch off a veritable monitor mania in the Union: Of the 84 ironclads constructed in the North throughout the Civil War, no less than 64 were of the monitor or turreted types. The first class of Union monitors were the 10 Catskills: Catskill, Camanche, Lehigh, Montauk, Nahant, Nantucket, Patapsco, Passaic, Sangamon, and Weehawken. (Camanche was shipped in knocked-down form to San Francisco. But the transporting vessel sank at the pier. Camanche was later salvaged, but the war was already over. Camanche thus has the distinction of being sunk before completion.) These ironclads, the first large armored warships to have more than two units built from the same plans, were awkwardly armed with one 11-inch and one 15-inch Dahlgren smoothbore. The Passaics were followed by the nine larger Canonicus class: Canonicus, Catawba (not completed in time for Union service), Mahopac, Manayunk, Manhattan, Oneonta, Saugus, Tecumseh, and Tippecanoe, distinguishable by their armament of two matching 15-inch smoothbores and the removal of the dangerous upper-deck overhang.

The eminent engineer James Eads designed four Milwaukee-class whaleback (sloping upper deck) double-turreted monitors: Chickasaw, Kickapoo, Milwaukee, and Winnebago. (Ericsson, on the other hand, loathed multiple-turret monitors, sarcastically comparing the arrangement to “two suns in the sky.”) Eads’s unique ironclads mounted two turrets, one of the Ericsson type (much to Ericsson’s disgust), the other of Eads’s own patented design: The guns’ recoil would actually drop the turret floor below the waterline for safe reloading; hydraulic power would then raise the floor back to the turret, wherein the guns could be run out by steam power. Eads’s two paddlewheel wooden-hull monitors, Osage and Neosho, designed for work on western rivers, were also unique. Although built to Eads’s designs, the two paddlewheel monitors mounted Ericsson turrets. All of the above monitors saw action in the U. S. Civil War. Completed too late for action were Marietta and Sandusky, iron-hulled river monitors constructed in Pittsburgh by the same firm that had built the U. S. Navy’s first iron ship, the paddle sloop USS Michigan.

Ericsson designed five supposedly oceangoing Union monitors: the iron-construction Dictator and Puritan, and the timber-built Agamenticus, Miantonomah, Monadnock, and Tonawanda.

The one-of-a-kind Union monitors were Roanoke, a cut-down wooden sloop; and Onondaga, also of timber-hull construction. Ozark, a wooden-hull light river monitor, had a higher freeboard than any Union monitor and also mounted a unique underwater gun of very questionable utility. None of the seagoing or the one-of-akind monitors saw combat.

Keokuk was an unlucky semimonitor (its two guns were mounted in two fixed armored towers and fired through three gun ports-a revolving turret would seem to have been an altogether simpler arrangement). The fatal flaw was in the armor, a respectable 5.75 inches, but it was alternated with wood. Participating in the U. S. Navy’s first attack on Charleston, South Carolina, Keokuk was riddled with some 90 Confederate shots and sank the next morning.

Aside from riverine/coastal ironclads, the Federals built only two broadside wooden ironclads, New Ironsides and Dunderberg (later Rochambeau, a super-New Ironsides, almost twice the former ironclad’s displacement), both with no particular design innovation. But New Ironsides could claim to be the most fired-upon ironclad during naval operations off Charleston, perhaps the most fired-upon warship of the nineteenth century, as well as the ironclad that, in turn, fired more rounds at the enemy than any other armored warship of the time. The broadside federal ironclad was formidably armed with fourteen 11-inch Dahlgren smoothbores and two 150-pound Parrott rifles, as well as a ram bow. Its standard 4.5-inch armor plate was far superior to the laminated plate of contemporary monitors. Whereas the monitors off Charleston suffered serious damage from Confederate batteries (and semimonitor Keokuk was sunk), New Ironsides could more or less brush off enemy projectiles and was put out of action only temporarily when attacked by a Confederate spar torpedo boat. During its unmatched 16-month tour of duty off Charleston, it proved a strong deterrent to any Confederate ironclad tempted to break the Union’s wooden blockading fleet off that port city, becoming the “guardian of the blockade.” Still, naval historians have tended to ignore New Ironsides and its wartime contributions because of the conservative design.

In light of their technological inferiority to British turret ironclads, it is difficult to understand why the Union’s Ericsson-turret monitors were also built by other countries: Brazil, Norway, Russia, and Sweden either built their own Ericsson-style monitors or had them built in other countries. (The Swedes, naturally enough, named their initial monitor John Ericsson.) The Russians constructed no less than ten Bronenosetz-class coast-defense monitors, and the Norwegians four similar Skorpionens. The Royal Navy ordered a class of four dwarf coastal ironclads that could be termed monitors, but they carried, of course, Coles turrets on breastworks well above the height at which they would have been mounted on Ericsson monitors, and they had superstructures. Furthermore, unlike the monitors, these coastal ironclads were in fact the diminutive template of the mastless turreted capital ship of the future.

The Union monitors, although an intriguing design, were in truth merely coastal and river warships; although several ventured onto the high seas, they only did so sealed up and unable to use their guns. Their extremely low freeboard (a long-armed man could have dipped his hand in the water from the deck) and tiny reserve of buoyancy made them liable to swamping, beginning with Monitor itself, which foundered off the North Carolina coast in December 1862. Monitor Tecumseh went down in less than two minutes after striking a mine at the Battle of Mobile Bay, the first instantaneous destruction of a warship, an all-too-common event in the twentieth century’s naval battles. Tecumseh was also the first ironclad to be sunk in battle, if one discounts two federal riverine armorclads sunk earlier at the Battle of Plumb Point Bend in May of 1862.

In fact, although the monitors might have been impervious to any Confederate gunnery, Southern mines destroyed the only three such warships sunk by the enemy: Patapsco, Tecumseh, and Milwaukee.(Monitor Weehawken foundered on a relatively calm sea in Charleston Harbor.)

The monitors also suffered from an extremely slow rate of fire; Monitor itself could get off only one shot about every seven minutes. Each shot required that the monitor’s turret revolve to where its floor ammunition hatch matched that of the hull; when firing, the two hatches were out of alignment to protect the magazine. And if an enemy shot hit where the turret met the upper deck, the turret could jam, something that apparently never happened to the many turrets built with Coles’s system.

In 1865, the U. S. Board of Ordnance obtusely argued that warships intended for sea service would be best with no armor at all. Yet at that very moment the Royal Navy had deployed five seagoing ironclads, including the magnificent pioneering Warrior and Black Prince, both warships with truly oceanic range, not to mention Defence, Resistance, and the timber-hull Royal Oak, Prince Consort, and Hector. The French, of course, years before had commissioned the seagoing La Gloire as well as Magenta and Solferino, the latter two the only ironclads ever to mount their main battery on double gun decks. (Magenta also has the melancholy distinction of being the first of the capital ships to be destroyed by mysterious explosion, a fate followed by about a score of such warships in the succeeding decades.)

In view of their design faults, plus their inferior and extremely slow firing guns and laminated armor, the monitors were a dead end in naval architecture from the start. The fact that Washington would consider the British sale of just two Coles turret rams to the Confederacy as grounds for war is a strong indication that the administration of President Abraham Lincoln realized the superiority of British-built turret ships to Union monitors.

Confederate Ironclads

Confederate secretary of the navy Stephen Mallory also wanted another type of ship for something far different from commerce raiding, one inspired by the old ship-of-the-line but possessed of some modern twists: an ironclad, steam-powered warship with rifled guns. He believed technological superiority would allow the South to overcome the disparity in numbers. “Such a vessel at this time could traverse the entire coast of the United States,” Mallory insisted, “prevent all blockades, and encounter, with a fair prospect of success, their entire navy.” They would allow the South to seize the naval initiative from its hidebound opponent. He eventually followed two routes to obtaining ironclads—buying them abroad and building them at home.

The Confederate Congress proved very receptive to Mallory’s ideas, voting $3 million to buy warships, including $2 million for ironclads. Mallory dispatched Lieutenant James North to Europe with instructions to try to buy a ship of the Gloire class, the innovative French ironclad commissioned in 1858. If this proved impossible, he should try to have one built. North, though, proved more interested in sightseeing than in doing his job. Mallory’s agents tried buying ironclads in Europe from May to July 1861, without success. The Confederate navy secretary decided to build them at home and signed deals for a few ships.26 Mallory also decided to build flotillas at various ports for their defense and gunboats for the Mississippi.

Building ironclads consumed most of the South’s naval effort. Mallory began studying the possibility of their construction in Southern yards in early June 1861. The first one arose from the burnt-out hulk of the USS Merrimack at Hampton Roads. The Confederacy had to do it this way because the South lacked the ability to build the ship it wanted from scratch. Mallory planned to use this new vessel, which became CSS Virginia, to clear the Union navy from Hampton Roads and Virginia’s ports. He generally believed that ironclad rams (which Virginia became) would be most useful for coastal defense. By late 1861, the Confederates had five ironclads in the works.

The Confederacy built ironclads to compensate for the enemy’s great numbers of warships. The South could not build oceangoing armored ships like Britain’s Warrior and France’s Gloire, but it could build slower, coastal ones like Virginia. These would, Mallory insisted, “enable us with a small number of vessels comparatively to keep our waters free from the enemy and ultimately to contest with them the possession of his own.” Mallory envisioned great but ultimately unrealistic achievements for Virginia. He believed that with a calm sea it could sail up the coast and attack New York City, causing such a panic that it would end the war. The Virginia’s success at Hampton Roads—ramming and sinking the USS Cumberland, then setting ablaze and driving aground the USS Congress—spurred Mallory to press the building of the CSS Louisiana in New Orleans, remarking that the “ship, if completed, would raise the blockade of every Gulf port in 10 days.”


Royal Navy Ships 1714–1815 I

It was for long an article of faith among naval historians that eighteenth-century British warships were inferior to their French and Spanish opponents, because British shipwrights remained wedded to craft traditions, while their Continental rivals were men of education who applied mathematics and science to the solution of their problems. This judgement flattered, and sometimes still flatters, a range of agreeable prejudices. It fitted the eighteenth-century upper classes’ admiration for France as the home of social glamour and prestige. It expressed British sea officers’ conviction that as men of honour they were both morally and practically superior to civilian technicians; it magnified their courage and judgement when they won, and excused their failures when they lost. It also increased their earnings when they were trying to sell their prizes to the Navy Board with a glowing endorsement of their virtues.

There are, nevertheless, several good reasons to reject the inferiority of British design out of hand. It is essentially an explanation of how France and Spain won the naval wars – which is not what we need to explain. In the century from 1714 more than half of all French warships (ships of the line and frigates) ended their careers sunk or captured, and the proportion rose steadily. In just over twenty years of warfare from 1793 to 1815, the French built 133 ships of the line and 127 frigates; and lost 112 and 126 respectively to enemy action or stress of weather. On average they lost a ship a month for twenty years. At first sight this does not suggest superior design. Moreover the comparison between ‘good’ French and ‘bad’ British design rests on the naive assumption that the two were directly comparable, that British and French designers were building the same size and types of ship, to fulfil the same functions – in other words that the strategic situations of the two countries were the same. This in fact is what many naval historians do assume: that the Bourbon powers, and subsequently Revolutionary and Imperial France, built their navies, and had to build their navies, to mount a frontal challenge to Britain for command of the sea, so that the opposing fleets may be considered as mirror images of one another. Command of the sea was the only thing worth striving for in the ‘Second Hundred Years’ War’, and Britain was the only enemy worth mentioning: in this view the historical function of the French and Spanish navies was to provide the Royal Navy with suitable opponents. These assumptions are extremely unsafe. As we have seen, there are good grounds for thinking that Maurepas and Patino were not planning to fight pitched battles with the British, and did not need ships designed for that purpose. The proper question to ask of all ship designs is not how well they compared with one another, but how well they corresponded to each country’s strategic priorities, and how wisely those priorities had been chosen.

Nor is it very useful to ask how ‘scientific’ the designs and designers of different countries were. It is still possible to encounter historians who put weight on the changing titles of the shipbuilders. In France ‘master carpenters’ (maîtres charpentiers) became ‘master constructors’ (maîtres constructeurs) and then simply ‘constructors’, before advancing to ‘constructor-engineers’ (ingénieurs-constructeurs) and finally becoming known as ‘naval architects’ (architectes navales), whereas in Britain warships were still being designed in the mid-nineteenth century by persons styled ‘master shipwrights’. The retention of a name drawn from the vulgar tongue, it is implied, must obviously indicate an unlettered craftsman confined to traditional rules, while a name derived from Latin must bespeak logic and education, and one based on Greek marks the summit of enlightened science. Perhaps it is still necessary to point out that the different titles of shipbuilders tell us something about their social aspirations, but nothing whatever about their working methods. Though British ship designers, like British professionals in comparable subjects such as architecture and engineering, continued to learn their business by apprenticeship until well into the nineteenth century, and though they were expected to spend a period working with their tools to understand the fundamentals of shipwrightry, the training they received in the mould lofts and drawing offices of the dockyards seems to have been in most respects as sophisticated as anything available in France.

There was, however, a real and important difference between Britain and France in attitudes towards ‘natural philosophy’, meaning science and fundamental knowledge in general. Mathematics lay at the heart of contemporary science, but mathematics was not an intellectually or socially neutral language. The mathematics of the ‘philosopher’ was pure mathematics: geometry, algebra, calculus. It was pure because it was abstract, and because it was essential to true science, that process of deriving universal truths from first principles, which Cartesianism prescribed. In social terms, this was the mathematics of the gentleman; one fully qualified for philosophy because he had no necessity to earn a living. It was very different from the vulgar utility of what in English was called ‘mixed mathematics’, the working calculations of men who had to work: men like bankers, tradesmen and navigators. The primacy of theory over practice, and of science over technology, was characteristic of France in the eighteenth century. The philosopher-mathematician alone was qualified to unravel the knottiest problems, and by tracing the fundamental machinery of nature he demonstrated his superior intellectual and social standing. ‘Tracing’ is the precise word, for geometry was a pure form of pure mathematics, and those whose subject could be expressed in geometrical terms enjoyed the highest scientific standing. It was a fundamental article of the Enlightenment faith that the philosopher was entitled and obliged to correct the work of the craftsman – this indeed was part of the official duties of the French Académie Royale des Sciences. As philosophers, gentlemen and mathematicians, its members were necessarily superior to mere practical experience. In naval architecture as in other domains, it was the duty of officers and philosophers to correct the vulgar errors of the shipwrights, by the application of pure mathematics.

The result was a series of studies by Leonhard Euler, Pierre Bouguer and others, deriving their prestige precisely from their remoteness from practical shipbuilding. The foundations they laid were built upon over the next two centuries to develop the modern science of naval architecture, but in the eighteenth century they had little to offer the shipwright. Most of their effort was devoted to the fashionable subject of hydrodynamics, and particularly the problem of the resistance of water to a moving hull, but since they ignored the existence of skin friction, which we now know to constitute virtually the whole of resistance at the speeds of which these ships were capable, their work had no practical value. More useful study was devoted to hydrostatics, which yielded the important definition of the metacentre, but French efforts to apply it in practice were not uniformly successful. The Scipion, Hercule and Pluton, launched at Rochefort in 1778 by Francois-Guillaume Clairin-Deslauriers, were among the first large French warships to have been designed on the basis of stability calculations. Unfortunately the sums were wrong, and the ships were too tender to carry sail. Much of their stowage had to be replaced by ballast before they could go to sea, sharply reducing their usefulness. Whatever else ‘science’ may have been doing in the eighteenth century, it was not an unmixed blessing to French naval architects.

One further general point about warship design needs to be made. Though ships may not have been directly comparable, naval architecture was highly competitive. Constructors constantly studied the designs of rivals at home and abroad, looking for ideas to borrow. In France and the Netherlands, so much less centralized in naval administration than Britain, these comparisons were often internal, between the rival traditions of the Mediterranean and Atlantic yards of France, and the admiralties of the United Provinces, but everywhere they were also international. All European navies were deeply involved in technical espionage, and in peacetime the French navy made a practice of sending its most talented constructors on extended visits to foreign, especially British, ports to learn everything they could. There is a particularly full and impressive report from the 1737 visit of Blaise Ollivier, master shipwright of Brest, with detailed comments on British and Dutch shipbuilding practice, much of which he admired and some of which he copied. All the European navies engaged in similar activities. In wartime they studied prizes; in peacetime they fished in the international market for warship designers. In 1727 the Admiralty of Amsterdam secured the services of three English shipwrights, with whose help it adopted ‘English-style’ designs – though naturally Rotterdam and Zealand declined to follow suit. In 1748 Ensenada, preparing to reform Spanish naval construction, sent Captain Jorge Juan on a major mission of industrial espionage to England. ‘His journey will be most useful to us,’ the minister wrote, ‘for in technical matters we are extremely ignorant, and what is worse, without realizing it.’ Juan returned with both information and a considerable number of shipwrights and artificers for the Spanish yards. English or Irish shipwrights became master shipwrights of Cadiz, Havana, Cartagena, Guarnizo and Ferrol. Throughout the eighteenth century the Danish navy, undoubtedly the world leader in technical intelligence, systematically collected copies of secret warship designs from every admiralty in Europe.

What seems to have been rare if not completely unknown in any navy was the literal copying of complete designs. Though statesmen and sea officers, impressed by foreign ships and ignorant of naval architecture, sometimes ordered ships to be built after the lines of a prize, it was in practice difficult if not impossible to do so. British hulls, for example, were more heavily timbered than French, so that a ship built in a British dockyard to the exact lines of a French design would displace more and float deeper. To maintain the same draught and freeboard, the British designer would have to adjust the lines, and so the ship would no longer be the same. In such cases the British designer might allow his superiors to believe that he had ‘copied’ a French design, or he might attempt to educate them in the complexities of naval architecture. Besides the lines, many other aspects of a foreign design would be changed to reflect British practice and requirements. The result might be a ship greatly influenced by foreign models, but it was never a slavish copy.

All these general considerations form a necessary background to any history of British warship design in the eighteenth century, but for thirty years, from the accession of George I in 1714 to the outbreak of war with France in 1744, British warships evolved slowly with little influence from outside. There were no Parliamentary votes for shipbuilding, so the practice of ’great rebuilds’ continued, though some of these ‘rebuilt’ ships were constructed years after their former selves had been broken up, in different dockyards, and without necessarily using any old timbers. The 1719 Establishment in principle dictated dimensions in detail, but in practice there seems to have been a slow but steady growth in dimensions. Some ships were built to the Admiralty’s 1733 proposal for a new and larger establishment, though it was never officially adopted, and the Ordnance Board blocked the heavier armaments which the Admiralty also wanted. Then the capture of the Spanish seventy-gun Princesa in April 1740, which took three British seventies six hours of hard fighting, caused considerable shock in Britain, and led to the adoption of a slightly larger 1741 Establishment, in conjunction with the heavier 1733 armament scheme. Soon afterwards the outbreak of the French war brought further shocks.

The main conservative forces affecting the British fleet were political and financial rather than technical. Neither the Navy nor Ordnance Board was enthusiastic about novelties which Parliament was not likely to favour, and still less likely to pay for. The Walpole administration kept up a large fleet, on paper and to a considerable extent in reality, and no one in the political world looked beyond numbers of ships to consider issues of quality and size. The Establishments were more an expression of this situation than an obstacle in themselves. The redoubtable Sir Jacob Acworth, Surveyor of the Navy from 1715 to 1749, did not take kindly to interference in the Navy Board’s business. ‘I have been in the Service fifty-seven years,’ he commented in April 1740 on complaints from Mathews,

and remember that the ships in King Charles’s time always decayed as fast, I am sure much faster, than they do now. But at that time, and long since, officers were glad to go to sea and would not suffer their ships to be complained of and torn to pieces in search for hidden defects.

The admirals resented it, and many would have agreed with Vernon that ‘the arbitrary power a half-experienced and half-judicious Surveyor of the Navy had been entrusted with had in my opinion half ruined the Navy’. But Acworth was no unthinking reactionary. He designed a number of ships whose underwater lines were based on theoretical concepts developed by Sir Isaac Newton. They were not a great success – a little more conservatism might have spared the Navy an unhelpful intervention from abstract science – but in many respects Acworth was a designer of talent. His ideas about the unhappy three-decker eighty-gun ships of the 1690s, and indeed about all the older British designs, stressed the importance of reducing topweight to improve stability and weatherliness. This was completely sound, and the admirals’ reactions may not have been unconnected with the fact that the tophamper Acworth wanted to remove consisted largely of their cabins.

The Bedford Admiralty arrived in December 1744 determined to reform British ship design together with everything else. Their chosen instrument of reform was a committee of senior officers under the chairmanship of Sir John Norris, directed to draw up a new establishment, and specifically to replace the three-decker eighties with a two-decker seventy-four-gun design. Although the committee consisted mainly of members of the Board or known opponents of Acworth, it was entirely dependent on him and the master shipwrights for technical advice, and proved as much a brake as a spur to progress. It moved some way in the direction of greater size, but flatly refused to abandon the small three-decker. The ships of the 1745 Establishment turned out to share many of the deficiencies of their predecessors: cramped, crank, overgunned and leewardly.

This impasse was broken by the sensation caused by the prizes of the two battles of Finisterre, above all the new French seventy-four Invincible. Maurepas’ new fleet was built around these seventy-four-gun two-deckers, with a lower-deck battery of twenty-eight 36-pounders and a main-deck battery of thirty 18-pounders. Though the Invincible was by no means the largest in her class (the Magnanime, taken next year, was considerably bigger) she was 50 per cent larger in tonnage than the standard British seventy-gun Third Rate, and fired a broadside 75 per cent heavier. The differences between these British and French ships arose almost entirely from the difference of size. Naval architecture is a question of balance: if two competent designers build rival ships of the same tonnage and type, one can only gain a marked advantage in any one quality, such as speed or armament, by sacrificing the others. Even a modest increase in size, however, permits a significant improvement in quality all round, and a 50 per cent increase ought to translate into overwhelming superiority. But increased size naturally means increased cost. British naval agitation to match or copy French designs was not so much a technical as a political campaign, directed at Parliament, to finance bigger and more expensive ships.

The Finisterre victories came too late in the war to have an immediate effect, but in 1750 the Sandwich Admiralty secured the Privy Council’s authorization to vary the 1745 Establishment as they thought fit, which effectively marks the end of the British shipbuilding establishments. In 1755, just as the outbreak of the Seven Years’ War released the Navy Estimates from peacetime financial limits, Anson was able to appoint a Surveyor of the Navy of his own mind, Sir Thomas Slade. From this date the Navy was in process of rapid transformation into a superficially French-style line of battle based on seventy-four-gun two-deckers.

There remained important differences, however, between British and French warships. British ships continued to be somewhat smaller in tonnage and shorter, but more heavily timbered and fastened. Their rig and lines performed best in going to windward, and in heavy weather. They were built to stand the strain of prolonged sea-time at all seasons, they were stored for long cruises, and they were built to fight. They were also built to last; relatively cheap to construct and maintain, they were the rational choice of a navy which meant to surpass its enemies both in numbers and in stamina. Their rig, masts, sails, cordage, blocks, pumps, cables, steering gear and fittings of every kind were greatly superior in design and quality. French ships of all classes were lightly built of inferior timber, fastened with nails instead of trenails, but their very long hulls were highly stressed in a seaway. In fine weather these ‘battle-cruisers’ with their long hulls and taunt rigs were fast off the wind, but their performance fell off rapidly when close hauled, or when wind and sea rose. What was worse French designers seem to have had something of an obsession with reducing the depth and weight of the hull, which made their ships light and buoyant, but directly weakened resistance to hogging, sagging and racking strains. Worst of all they actually believed that the working of the timbers increased the speed of the ship. Consequently these ships had high building costs, high maintenance costs and short working lives, which made France’s low investment in docks and yards all the more expensive. In close action French ships with their light scantlings were a death trap.

Some French observers were aware of some of the deficiencies of their designs. A warship, one constructor declared,

ought to be fast, so everything is normally sacrificed to that. They are lightly timbered in order to be buoyant and carry their guns high; they have fewer and weaker fastenings because the play of the timbers makes for speed… it is to be feared that these principles lead the king’s constructors to build ships of the line which lack some of the qualities of a real man-of-war. They are afraid of losing their reputations, because the height of success for them is a fast ship which carries her guns high.

French dockyard officers had to pick up the pieces, literally. The constructors, complained the comte de Roquefeuil, commanding at Brest in 1771,

are all frauds. They build ships which are very light, very long and very weakly fastened because they sacrifice everything to speed and that is the way to get it. The first cruise gives the ship and her builder their reputation… [afterwards] we have to rebuild them here at great expense for a second commission by which time they have lost their boasted speed.

It is not even clear that sacrificing so much to hull forms which were fast in certain circumstances was actually the best way to get high speed in practice. Modern studies suggest that the possible differences in hull form, within the inherent limitations of wooden ship construction, cannot account for the wide differences in recorded performance. The smoothness of the underwater hull, which was a matter of cleaning or coppering (and hence of docks), and the infinite variations of rig and trim which were under the captain’s control, were almost certainly worth more. This explains how frequently British ships were able to catch French ones even in conditions which should have favoured them, and why French prizes taken into British service seem generally to have been faster after capture than before. Moreover prizes were usually significantly altered. The ships were always rerigged and rearmed, and the holds (especially of frigates) were rebuilt to give increased stowage to allow for prolonged cruising. The hanging of the decks, the siting of hatchways and magazines, the stowage of boats and booms, the position and design of pumps and capstans were often changed. These alterations produced substantially different ships.

Mention of frigates calls us back to the other important innovation in mid-eighteenth-century warship design. The new French battle-fleet of two-decker sixty-fours, seventy-fours and eighties were unquestionably built to a common plan imposed from Versailles, though the actual hull designs differed from yard to yard. Small cruisers, however, were beneath the minister’s notice, and the constructors were left to build more or less what they thought fit. It seems therefore that the Médée of 1740, commonly regarded as the first of the ‘true’ or classic frigate type which formed so prominent a part of all navies by the late eighteenth century, was a product of Ollivier’s unaided genius. The essence of the frigate in this sense was a small two-decker cruising warship mounting no guns on the lower deck. This made it possible to carry a battery of relatively heavy pieces on the main deck, high above the waterline, where they could be fought in bad weather, as well as lighter guns on quarterdeck and forecastle. This general arrangement was not new, in French or British service. In 1689 Torrington had proposed

that these new frigates should for rendering them more useful for their Majesty’s service, be built in such a manner that they should have but one size of ordnance flush, and that to be upon the upper deck, whereby they will be able to carry them out in all weathers.

The resulting class of Fifth Rates were soon overloaded with guns, in the British style. They were succeeded in 1719 by a class of Sixth Rates carrying a battery of twenty 6-pounders (ten ports a side) in the same arrangement, but they too tended to become overloaded with guns as British officers watched with concern the growth in the power and size of foreign warships. When the outbreak of war with France in 1744 exposed British trade to attack by French warships and privateers, the small, slow and cramped British cruisers aroused widespread dissatisfaction in the Navy. Once again it was French prizes which provided the leverage to dislodge the Navy Board’s opposition. ‘As all our frigates sail wretchedly,’ Anson wrote to Bedford in April 1747,

I entreat your Grace that an order may be immediately sent from your Board to the Navy Board to direct Mr Slade the Builder at Plymouth to take off the body of the French Tyger with the utmost exactness, and that two frigates may be ordered to be built with all possible dispatch, of her dimensions and as similar to her as the builder’s art will allow; let Slade have the building of one of them.

The Navy Board mounted a stout defence of the small forty-gun two-decker as superior to French cruisers like the privateer Tigre to which Anson referred, and they had some grounds to do so, for the French designs had all the characteristic French weaknesses, being very long and flimsy. When Ollivier’s Médée was taken in 1744 the Navy Board refused to buy her, so she was sold as a privateer, and soon afterwards fell apart in the open sea. This was not what the Navy wanted, and in spite of Anson’s request for exact copies of a prize, this was not what it got. The visible superiority of French cruisers, at least in speed, provided the Bedford Board with the leverage it needed to overcome the Navy Board’s resistance, and the co-operation of dockyard shipwrights of a younger generation than Acworth provided the technical backing – but what they produced were not exact copies of French designs. It is clear from the surviving correspondence that during the 1740s the shipwrights were carefully comparing prizes with the fastest existing British designs (notably the yacht Royal Caroline), and using the untutored enthusiasm of Anson and his colleagues to back a move from the old short two-deckers to the first British ‘true frigates’, with longer hulls (eleven or twelve ports a side) giving a twenty-two or twenty-four-gun battery, initially of 9-pounders. These very successful ships were inspired by French prizes in a political as much as a technical sense. The differences of British from French design philosophy and performance were even clearer in the case of frigates than of ships of the line.

Using the standard shorthand method by which all navies classified their ships by the number of guns, the early British frigate classes were mostly twenty-eights, which were followed in the Seven Years’ War by thirty-twos. In the case of frigates, however, the number of guns is not a good measure, partly because it included the light guns on quarterdeck and forecastle which could easily be changed, but mainly because it concealed the most important factor, the calibre of the main battery. Though a thirty-two does not sound much more powerful than a twenty-eight, the twenty-eights had a 9-pounder main armament and the thirty-twos, 12-pounders, giving a broadside 50 per cent heavier. These ships in turn were followed in the American War by the first 18-pounder frigates, rated as thirty-sixes or thirty-eights, but with more than double the broadside of the twenty-eights. It is therefore most useful to refer to frigates, as many contemporaries did, by their main battery calibre, and especially to distinguish the 18-pounder ‘heavy frigates’ from their predecessors.

The smaller cruisers known as sloops developed in parallel with the frigate, of which they were essentially miniature versions with one less deck, carrying their battery on the open upper deck. By the time of the American War many of the smaller sloops were rigged as brigs rather than ships. The two-masted rig economized significantly in manpower and was perfectly satisfactory for most purposes, though more vulnerable to damage in action. As an alternative it was possible to rig vessels of this size (200 tons or so) as cutters or schooners, which were even more economical in manpower, but whose very big sails required expert handling. Smallest of all sea-going warships were the gunbrigs and gunboats, built in considerable numbers for Channel patrols and local defence during the Great Wars.