Japanese Siege Weapons

Early Fortifications

Although fortifications were constructed in Japan prior to the feudal period, frequent conflicts associated with warrior ascendancy inspired new, distinctive temporary architectural forms as well as more lasting structures to protect against military attack.

Up to the beginning of the feudal era, three forms of fortifications were built, according to archaeologists. The grid-pattern city form was inspired by Chinese planning precedents, and included gates or walled enclosures. Mountain fortresses appear to be an indigenous form, and were typical of remote areas. Plateaus or plains often utilized the palisade, a semi-permanent defense. Typical defenses included a rampart, a ditch, and a palisade. Grid-pattern cities were surrounded by walls that served as a demarcation point rather than as true protection, and eventually such barriers disappeared. Remains of mountain fortresses found in northern Kyushu were a more effective means of protection, and may have belonged to ancient kingdoms that ruled parts of Japan in early times. Palisades were often constructed in the northeastern areas of the main island of Honshu. Although excavations have revealed only partial remains of such structures, they are significant since they offer prototypes for medieval fortifications.

Until the end of the Kamakura period, most fortresses built in Japan were relatively simple, and were designed for a particular siege or campaign. Terms such as shiro and jokaku (translated in later eras as “castle”) appear frequently in 12th- and 13thcentury accounts of warfare, but in the Kamakura era, these terms refer to temporary fortifications. Early medieval defense structures were more like barricades than buildings, and were not intended to house soldiers for extended periods. However, such fortifications could be elaborate and large in scale.

Many extant screens depict scenes from Tale of the Heike, the famous Japanese historical saga of the twelfth and thirteenth centuries. Its twelve volumes relate the long narrative of the rise and fall of two rival warrior clans, the Genji (or Minamoto) and the Heike (or Taira). This screen illustrates famous episodes from the saga, including the battle at Ichi-no-tani Mountain and the death of Taira no Atsumori at Yashima.

Literary and pictorial accounts confirm that extensive planning and earthworks projects were utilized throughout the medieval era for major battles. For instance, the defense works at Ichinotani erected by the Taira clan in 1184 included boulders topped by thick logs, a double row of shields, and turrets with openings for shooting. Even if descriptions of such structures taken from accounts of the Gempei War dating to the late Kamakura era exaggerate these defenses, they capture the labor, time, and ingenuity involved in such efforts.

As wartime construction continued, Japanese military architects became skilled in adapting civilian structures that offered multiple options for warrior defenses. Composite barriers utilizing timber and other materials that protected crops from intruders and animals were helpful in subduing infantry offenses. Military architects familiar with agricultural irrigation principles constructed ditches and moats to deter mounted troops. In sum, military construction of the early medieval period involved tailoring familiar forms to warrior needs to provide an initial line of defense.

Some temporary construction types afforded flexibility and served well in both offensive and defensive situations. Kaidate (shield walls) and sakamogi (brush barricades; literally “stacked wood”) were both in common use by the 13th century. Kaidate, formed of rows of standing shields, had been employed since the end of the Asuka period (eighth century), and were valuable as portable field fortifications. Sakamogi, which were most likely inspired by barriers for livestock, were useful in several contexts as well. These deceptively simple structures continued to be effective in the age of gunpowder as they remained difficult to cross and also resisted explosive shells. Barriers made of shields could be made more effective through deployment atop, or in front of, another defensive form. However, as the power of the Ashikaga shoguns declined in the Northern and Southern Courts era, combat conditions changed, and samurai clans confronted elevated fortresses where warriors on horseback were ineffective.

Azuchi-Momoyama- and Edo-Period Castles

After the feudal system was reorganized by the Tokugawa shogunate, castles (shiro) were erected in the center of a daimyo’s domain, so they would be easily accessible. Without natural defenses such as hills and plateaus, these structures required additional protection compared with the elevated shiro built during the late Muromachi and Momoyama periods. For security, builders developed walls of enormous boulders that often had smooth surfaces that would be difficult to scale. Moats (hori) also provided a means to deter an attacking force.

The castle was not only a means of defense, but also served as the hub of administration and commerce in the domain. Castles housed the domain lord and chief retainers. Towns developed around the structures, called “towns beneath the castle” (jokamachi) since the castle was often elevated, and both literally and figuratively overshadowed all other buildings nearby. Merchants and artisans became an important aspect of life in these castle communities, as daimyo and their retainers had more time and disposable income than in the past. Further, the rise of fashion and interest in display (in the sense of decoration and adornment) that arose in the cosmopolitan Edo period made it necessary for members of the warrior class to keep up appearances, and this led to healthy economic growth even in provincial castle towns.


In the late Kamakura and early Muromachi eras, locally powerful landholders did not yet have the resources to commission and train considerable numbers of mounted warriors. At this stage, extensive forces designed for long-distance campaigns were unnecessary as well, since battles in the provinces often culminated in localized sieges to gain control of a strategically positioned castle. Thus significant numbers of well-trained foot soldiers were necessary to enter the territory of an opponent and scale his fortress. While swordsmanship began to gain prominence among samurai skills in the early medieval era, the primary warrior weapon among foot soldiers was a long polemounted arm called naginata, and this was supplemented by archery. Military drills using polearms involved learning to pull a cavalryman from his mount and engage him in close-range combat. Other practical applications of such weapons included thrusting, or throwing, a spear or other polearms in order to hit a distant target. Archers and infantry equipped with spears were also trained to send arrows over castle walls to cover the approach of foot soldiers who sought to scale the walls and thereby gain access to the castle.


Siege Warfare

Japanese castles were adapted to siege warfare. Again the similarities between Western and Eastern warfare are evident and the sophistication of Japanese siege-craft is obvious.

If a samurai is on the defending force, the following items are things he would be familiar with as he moved in and around the castle.

Arrow and Gun Ports

Inset into the walls of castle defenses are small holes—they are normally rectangular, circular or triangular. Positioned at different heights, the defenders use them to shoot out over the field of battle. However, shinobi creep up to these apertures and fire burning arrows and flash arrows through them into the interior of the castle grounds to discover details about the interior layout. In addition to this, they would throw in hand grenades to kill those shooting out at the opposition.

Stanchions, Walkways and Shields

Along the inside of castle walls, wooden stanchions and frames would support multiple levels of walkways—similar to modern day scaffolding. Samurai would use these levels from which to shoot outward, either through arrow and gun ports or over the tops of castle battlements from between shields. In addition to this, bridges that could be retracted were set up at various positions; if the enemy breached the defenses these walkways could be retracted, allowing defending samurai to kill the enemy from the opposite side.

Killing Zones

Walls, turrets and enclosures were created to form killing grounds and zones, where the defending army could attack the enemy with crossfire and pin them into a corner and halt movement.

Turrets and Palisades

As discussed before, the castles of the early Sengoku Period and before were generally smaller; walls could be protected by turret towers, wooden shields and semipermanent buildings that were made of wood and were built along the tops of walls. Shinobi had various mixtures that would set fire to these, fires that would be difficult to extinguish, helping to break through the castle defenses.

Allied Help

A defending castle could set up a series of fire beacons and send messenger relays to request allied forces to counter the siege. Sometimes the relieving force could surround the besiegers, forcing them to defend their own rear and fight on two fronts.

Sallies and Sorties

The castle would send out night raids and attacks when they thought that the time was right. They may even evacuate a castle from an non-besieged section—if any—through gates and ports. Shinobi were trained to watch the smoke rising from castles. If the smoke from cooking fires and kitchens was too much, too little, or later than normal, a shinobi would know that the enemy had either started to evacuate or that they were preparing extra food for those going on night raids or that the food stores were diminishing. All of which was information the shinobi would pass on to his commander.

Those who were attacking the castle had certain weapons and tools to help degrade the height and protection advantages of the defenders.

Trench Warfare

Trenches at their smallest were three feet deep with an earth mound on the top of around two feet; this total of five feet covered the average height of a samurai. The closer to the castle the trench lines were the deeper they had to be dug, as arrows could be shot into the defenses from such an angle.

Towers and Constructed Turrets

As discussed previously, the enemy battle camp had collapsible turrets and towers; these were erected to see enemy troop movements and shinobi.

Battering Rams

Covered rams on wheels were used to take down castle doors and break open sections of defenses.

Shields and Walls on Wheels

Small platforms were placed on low carts with walls erected on the front. These walls had shooting ports and would be rolled into place, and from here attacking samurai could shoot at the enemy. This included walls mounted on arms that could be raised so that samurai could shoot out from below them and other such contraptions.

Shields, Bamboo Fences and Bundles

Human-sized wooden shields that stood erect with the help of a hinged single leg would protect samurai. In addition to this, bamboo was tied in large bundles and shooting ports were cut out of the middle. These bundles could be leaned against waist-height temporary fences so that samurai could shoot from behind cover.

Cannon and Fire

Cannon were used to launch fire and incendiary weapons and shot. Kajutsu—“the skills of fire”—included long-range rockets, flares and anything that causes flames in the enemy camp. Some shinobi were essentially agents who moved into the enemy castle and made sure that fires were set from within. One shinobi trick was to set a fire away from the main target to distract the defenders from the actual target and then to move on with their initial aim of setting fire to more important things like the main compound.


Tunneling was undertaken to undermine the enemy defenses. If done in secret and not on a war front, the tunnel had to start far from the target, or start from inside a nearby house. To discover if tunneling was taking place, empty barrels would be set into the ground to listen for mining below.

Moat Crossing Skills

Portable bridges and temporary structures were used to cross rivers and bridges. The shinobi’s task was to discover the length, width and depth of a moat and report the dimensions, or to cross it in secret at night.

On the whole, the samurai castle was a place of residence and the target of a siege. The samurai would defend and attack castles with ingenious tricks and tactics and shinobi on both sides would come and go, stealing information or setting fires to things, something that was quite normal in life as a samurai.


Gului-gorod and Lines of Defence

Gului-gorod or Gulyay-gorod

Bereg Line – Pre-Ivan IV

1.    Nizhni Novgorod to Zvenigorod Line (largely riverine line along Moskva river not itself shown)

2     Kolomna to Kaluga Line (along Oka)

Ivan IV (1547–1584)

3.    Novgorod Severski (Desna) Line – Alatir (Sura River) to Putivl’

4.    Zasechnaya Line – Pereslavl-Ryazanski (Oka) to Kozelisk (Zhizdra)

5.    Zasechnaya Line – Alatir to Skopin (Don)

Boris Godunov (1585–1605)

6.    Kromi (Oka) to Elets (near Don)

7.    Kursk (Seum) to Voronezh (Don)

Belgorodskaya Cherta (Seventeenth Century)

8.    Voronezh-Tsna Corridor (1635)

9.    Belgorodskaya Cherta (1635-1646) (Tambov (Tsna) to Akhtyrka (Vorskla))

10.  Simbirskaya Cherta (1648-1684) (Tambov to Simbursk (Volga))

11.  Zakamskaya Cherta (1652-1656) (Simbursk to Meznelunsk (Kama))

12.  Seranskaya Cherta (1683–1684) (Simbirskaya Cherta via Penza to Sizran (Volga))

Izium Cherta

13.  (1680s) Userdsk (Belgorod Line) via Izium to Kolomak River)

Between the fifteenth and seventeenth centuries the Russian army developed linear barriers called Gului-gorod or walking towns. These were intended for operation on the steppe where they could provide cover from nomad attacks and arrows, and moveable palisades through which firearms could shoot down mounted nomads. These were unlike earlier war wagons, where men fought from, on and between the vehicles, as they consisted of upright rectangular wooden shields, which moved on bogies or runners and which had gun ports. Before battle was joined, the units of the Gului-gorod were coupled together like railway carriages to form a continuous linear barrier. The men stood directly on the ground and there were no gaps between the individual units. Thus something that genuinely looked like a wall was presented to the enemy. Their value eventually diminished when nomads brought with them their own field artillery – as did the Crimean Tatars – but in the sixteenth century they played an important part in the shift of the Russian state from the defensive to the offensive. They were widely used in the sixteenth century wars with the Kazan and later by the Ukrainian Cossacks.

The Battle of Molodi in 1572, during the reign of Ivan the Terrible (1547–1584), perhaps marked a turning point in the fight between settled states and nomads. In 1571 Khan Dovlet Girai of the Crimean Khanate had burned Moscow; he reputedly killed 100,000 people and enslaved another 150,000. The following year Russians used the Gului-gorod to shelter their artillery and harquebusiers from Tatar archery. After several days of fighting a detachment of Russians left the protection of the Gului-gorod and attacked the Tatars in the rear while those who had remained behind the Gului-gorod advanced before it, crushing the Tatars. Dovlet Girai’s son, grandson and son-in-law were killed and only 20,000 of an army of 150,000 Tatars returned. The Ottoman Turks had supported their Tatar vassals and the Russian victory at Molodi, although little known in the west, was seen as a key to reversing the northern expansion of the Turks. A crucial role in the victory was accorded to the mobile linear barriers or Gului-gorod.

Zasechnaya Cherta – The Zasechnaya System was built during the reigns of Ivan IV, the Terrible, and his able servant, and then successor, Boris Godunov, who served as Regent from 1585 to 1598, and then Tsar, from 1598 to 1605. (Sometimes the system is called the Bolishnya Zasechnaya Cherta or Great Abatis Line.)

A line ran from Alatir on the Sura River, in the east, via Shatsk to Novgorod Seversky, on the Desna, to the west. Shatsk blocked the Nogai Shlyah.

The main construction effort was the Zasechnaya Cherta which consisted of two continuous lines joined by the Don River. The more westerly line ran from Pereslavl-Ryazanski via Tula (where the Muravsky Shlyah was blocked) to Kozelisk. To the east, and taking up the earlier line from Alatir to Shatsk, the continuous line finished at Skopin on the Don to the west of Shatsk.

Under Boris Godunov further shorter lines appear to have been built across the Muravsky Shlyah between Kromi (River Oka) and Elets (near the Don) and to the southern towns of Kursk (River Seym) and Voronezh (Don).

Belgorodskaya Cherta – This line started as a short barrier blocking the corridor that stretched from Voronezh to Tsna. The line extended over the entire southern frontier.

Kozlov Steppe Wall – Voronezh-Tsna Corridor – As Russian estates and peoples expanded south of the Zasechnaya Cherta they remained vulnerable to the Nogai Tatars who were moving up the Nogai Shlyah. So a decision was made in the mid-1630s to build a twenty-five kilometre earthwork to block the open gap in the Voronezh-Tsna corridor across the Nogai Shlyah between the Polnoi Voroneszh, a tributary of the Voronezh, and the Chelnovaia River, a tributary of the Tsna. Initially a palisade was envisaged, but the risk of its being burnt was considered too great and a linear earthwork was constructed. This was designed by a Dutch engineer, Jan Cornelius von Rodenburg to incorporate both cannon and handgun positions. ‘The earth steppe wall began at Belyi Kolodez’ on the right bank of the Pol’noi Voronezh River, sixteen kilometres west of Kozlov, and ran eastward for twenty-five kilometres until it reached the left bank of the Chelnovaia River.’ To the east of the Chelnovaia was an abatis barrier which filled the gap to Tambov on the Tsna River. The barrier was built on time and budget and it worked immediately. In 1636 a force of 10,000 Tatars was driven off. The success of the Kozlov defences convinced the Russians that it was possible to move the main frontier down from the Zasechnaya Cherta (Abatis Line) to a new line running west from a wall on the Kozlov steppe, so as to cross all the shlyahs.

Belgorodskaya Cherta – This ran along the southern edge of forest steppe from, at the west end, Akhtyrka on the Vorskla, and east to Tambov on the Tsna, for 800 kilometres, of which 140 kilometres consisted of earthworks.

Simbirskaya Cherta – The Belgorodskaya Cherta was continued by the Simbirskaya Cherta from Tambov on the Tsna, to Saransk, west of the Sura, on towards Tagay, between the Sura and the Sviyaga Rivers, and then to Simbirsk on the Volga totalling 500 kilometres. It was completed in 1658.

Seranskaya Cherta – Roughly half way between Tambov and Simbursk, the Seranskaya Cherta split off from the Simbirskaya Cherta and ran about 300 kilometres past Penza to Sizran on the Volga.

Zakamskaya Cherta – The move by Russians eastwards disrupted the Turkic Bashkirs and so further linear barriers were constructed that ran east of the Volga. The Zakamskaya Cherta went east of Simbursk on the Volga, to the Meznelunsk on the Kama, for about 400 kilometres.

Izium Line – In 1679–1680 most of the steppe along the Donets and Oskol rivers was enclosed behind yet another new more southerly barrier, the Izium Line. The enclosed area formed a triangle whose base was the Belgorod Line. The Line ran along the lengths of the Oskol and the Donets Rivers, and at their confluence stood Izium as the point of the triangle. Whereas the eastern end diverged from the Belgorod Line near Userdsk, the western end did not quite return to the Belgorod Line but stopped on the Kolomak River. (Kolomak is about sixty kilometres south-west of the west end of the Belgorod Line at Akhtyrka.) The Izium Line brought the Russian lines another 160 kilometres south-east of the Belgorod Line and about 150 kilometres from the Black Sea. The line itself was over 500 kilometres long. The area inside the line corresponds approximately to Kharkiv Oblast and contains the modern city of Kharkiv which began as a small fort in about 1630, and today is the second largest city in the independent Republic of Ukraine.

Pskov-Smolensk-Briansk line – A rare example of a linear barrier which was built by one state against another was seen when Peter the Great ordered the construction of an abatis against Charles XII of Sweden. From 1706 to 1708 a fortified western border was established along the Pskov-Smolensk-Briansk line as a desperate measure. In 1724, Peter I introduced the fortress system of defending borders (primarily in the west), but Russia continued to use frontier lines in the south and east.

The lines built by Muscovy and then Russia worked on a similar basis to the Kievan Rus’ Zmievi Vali: ‘The object of all these linear defensive systems was impede the advance of the Tatars, restrict the mobility of cavalry, and gain time to allow the civilian population to be evacuated and large forces to be summoned from nearby fortresses. The Tatars had no siege machinery or great knowledge of siege warfare, so unsophisticated earth and earth-and-timber fortifications served the purpose well, and were cheap to build.’

By the admission of the Tatars they were successful: ‘Even the Tatars themselves acknowledged this (fortification had blocked the Nogai Shlyah), some Tatar princes … admitting to Russian interpreters in 1637 that the new defences at Kozlov and Tambov had ‘shut down’ the Nogai Road.’ Also: ‘A 1681 report of the Military Chancellery to the Boyer Duma assessed the impact of the Belgorod Line in the most enthusiastic terms, claiming not only that ‘enemy warriors can no longer attack the Borderland towns by surprise because of that defence line,’ but that military colonisation along the Line had even transformed the national economy by making the southern frontier safe for private colonisation.’ Previously peasants had preferred to risk their chances in Siberia rather than be taken slave by the Tatars. Now they stayed put and opened up the extremely fertile black earth belt. The population of the Moscovite state doubled in the quarter century after the construction of the Belgorod Line.

With the fall of Kiev the heartland of Russia moved to the less hospitable but more defendable north – with Muscovy emerging as the dominant state. It then developed technology involving firearms, linear barriers both moveable (like the Gului-gorod) and static, in addition to infantry and cavalry. Using these means Muscovy managed to stabilise the southern frontier and then to move it further south, displacing the nomads in the process. In the eighteenth century, it was the nomads whose own linear defences were stormed by the Russians when the Perekop Wall across the Crimean isthmus was breached. In Russia, as this chapter has hopefully shown, linear barriers played a crucial and aggressive part in the grand exercise of political and military expansion.


Martello Tower

Fully restored Martello Tower (No.7) in Killiney Bay, Co. Dublin, Ireland.

The British were never shy about adapting foreign military designs. The Martello towers were a copy of the round fortress at Mortella (Myrtle) Point in Corsica, designed by Giovan Giacomo Paleari Fratino and completed in 1565.

The Corsicans had previously built similar towers at strategic points to protect themselves from pirates. They were two storeys high and measured 12–15m (36–45ft) in diameter, with a single doorway 5m off the ground accessible only by a removable ladder. Watchmen, employed by local villagers, would signal the approach of suspected threats by lighting a rooftop beacon which would alert the local defence forces to the threat.

On 7 February 1794, the British warships HMS Fortitude and HMS Juno unsuccessfully attacked the Mortello tower and were repelled. Only heavy fighting by land troops took the town below the tower. Vice-Admiral Lord Hood reported:

…The Fortitude and Juno were ordered against it, without making the least impression by a continued cannonade of two hours and a half; and the former ship being very much damaged by red-hot shot, both hauled off. The walls of the Tower were of a prodigious thickness, and the parapet, where there were two eighteen-pounders, was lined with bass junk, 5ft from the walls, and filled up with sand; and although it was cannonaded from the Height for two days, within 150yds, and appeared in a very shattered state, the enemy still held out; but a few hot shot setting fire to the bass, made them call for quarter. The number of men in the Tower were 33; only two were wounded, and those mortally.

The British were hugely impressed by the tower’s effectiveness and copied the design, misspelling the name. When the British withdrew from Corsica in 1803, with great difficulty they blew up the tower.

The classic British Martello tower, built to repel a Napoleonic invasion, consisted of three storeys above a basement. The ground floor held the magazine and storerooms, the first floor housed 24 men and one officer, and had fireplaces built into the walls for cooking and heating. The flat roof was surmounted with one or two cannon on a central pivot that enabled the guns to rotate 360 degrees. A well or cistern within the fort supplied the garrison with water. An internal drainage system linked to the roof enabled rainwater to refill the cistern.

Around 140 were initially built, mostly along England’s south coast. Great Britain and Ireland, united as a single political entity in 1801 to 1922, used Martellos as a defensive screen covering both main islands. Chains of Martello towers were built.

Between 1804 and 1812 the British authorities built 103 in England, at regular intervals from Seaford in Sussex to Aldeburgh in Suffolk. Most were constructed under the direction of General William Twiss (1745–1827) and a Captain Ford. There were also three much larger circular forts or redoubts that were constructed at Harwich, Dymchurch and Eastbourne which acted as supply depots for the smaller towers as well as being powerful fortifications in their own right. None were tested in combat against Napoleon’s forces, but they were effective in combating smuggling.

After the threat had passed, of the Martello towers in England, 15 were demolished for raw materials, 30 were washed away by the sea, and four were reduced to rubble as part of an exercise to test the power of new artillery. During the Second World War many were used as observation posts and as anti-aircraft gun platforms. Of the remaining 47, some have become museums and galleries, others distinctive private homes, while a few are derelict.

Three Martello towers were built in Scotland, two at Hackness and Crockness the Orkney Islands, between 1813 and 1815 to guard against the threat of French and American raiders attacking convoys assembling offshore. A third Scottish tower was built on offshore rocks facing the Firth of Forth in 1807–09 to defend Leith Harbour.

A small number of Martello towers were also built in Wales but few survived. The most notable exceptions are in Pembrokeshire Dock.

Around 50 martellos were built around the Irish coastline, from Drogheda, to Bray on the easy coast, around Dublin Bay and around Cork Harbour. The most famous, but for non-martial reasons, is that along the coast from Dublin at Sandycove. James Joyce lived there for a few days with the surgeon, politician and writer Oliver St John Gogarty. In Ulysses, Joyce’s fictional character Stephen Dedalus lives in the tower with a medical student, Malachi ‘Buck’ Mulligan. During the 1980s, the rock star Bono owned the Martello in Bray, County Wicklow.

Martellos were built across the Channel Islands. Three, built in 1804, are in Guernsey: Forts Grey, Hommet and Saumarez. The oval Brehon Tower, built in 1856, represents the final evolution of the Martello tower.

Eight were built in Jersey, three between 1808 and 1810 and five between 1834 and 1837, one of which, L’Etacq, the German occupation forces destroyed during the Second World War. The three original towers are the Tower, Portelet, and la Tour de Vinde. The four surviving, later towers are Lewis’s and Kempt, both in St Ouen, Victoria and La Collette. In addition, there are the Jersey Round Towers and the Guernsey loophole towers, often called Martellos, built in the late eighteenth century as precursors.

Martellos were copied across the globe until the 1870s, although by then it had long been clear that they could not withstand the new generation of rifled artillery.

The French themselves built similar towers along their own coastline as platforms for optical telegraphs. The United States and the Dutch also built a number. But the greatest number remained British-built, protecting her swiftly expanding empire.

During the British occupation (1798–1802) of Minorca, Governor Sir Charles Stuart ordered Engineer Captain Robert D′Arcy to build 12 Martello towers along the coast. These, when added to the three Spanish towers already in place, gave Minorca 15 towers. One tower, the Princess Tower, or the Erskine Tower, was incorporated into the Fortress of Isabelli, built between 1850 and 1875. The tower was converted to a powder magazine, but when lightning struck, the subsequent explosion destroyed it.

In Minorca the British erected Stuart’s Tower in 1798 on Hangman’s Hill at San Estaban or Saint Stephen’s Bay. In 1756 and again in 1781, batteries on the hill had supported successful attacks on the Fortress. The tower was built both to secure the hill and protect the entrance to the bay. To protect the harbour of Fornells, a tower was built on the rocky headland overlooking the harbour’s mouth, and a small tower on the island of Sargantana. Finally. Another was built at Santandria to protect the old capital of Ciudadela.

During the British ‘Protectorate’ of Sicily after the escape of the Bourbons from Naples, Sicily began to build towers to resist an invasion by Napoleon’s armies. The new higher rate of fire of ships’ guns led to the choice of the Martello tower as the model. The Sicilian Martello towers were built around 1810. One was the Magnisi tower at Syracuse, later used by the Italian Navy as an observation post during the Second World War.

On Bermuda, Martellos were built behind Ferry Island fort and at Ferry Reach. The latter was built of Bermuda limestone from 1822–1823, with walls up to 11ft thick and surrounded by a dry moat. Its role was to impede any attack on St George’s Island from the main island of Bermuda. Two more Martello towers were planned, but never built, to protect the dockyard.

On Jamaica the Spanish slave agent had in 1709 built a fort in Harbour View, to guard his home against attack. The later English Governor, George Nugent, later strengthened the fort to guard the eastern entrance of the city of Kingston Harbour. Fort Nugent was built between 1808 and 1811 at a cost of some £12,000.

Fourteen Martellos were built in Canada, nine of which remain. The Canadian Martellos were fitted with removable cone-shaped roofs to protect against snow. Halifax, Nova Scotia, had five. The oldest, the Prince of Wales tower in Point Pleasant Park, is also the oldest of that style in North America. Built in 1796, it was used as a redoubt and a powder magazine. Quebec originally had four Martellos. Tower No. 1 is on the Plains of Abraham, overlooking the St Lawrence River, scene of Wolfe’s earlier victory. Four more were built at Kingston, Ontario, to defend its harbour and naval shipyards. Fort Frederick was constructed with elaborate defences, including earthen ramparts and a limestone curtain wall. The Shoal tower is the only one completely surrounded by water.

In April 2006 that the Canadian military named a Forward Operating Base in Afghanistan FOB Martello.

The British built five Martello towers in Mauritius between 1832 and 1835 at Grand River North West, Black River and Port Louis, of which three survive.

A Martello was built at Freetown, Sierra Leone, in 1805 to defend the port from attacks by the warlike Temne tribe, one of the two dominant ethnic groups there. It was significantly modified in 1870 when it was truncated to allow the installation of a water tank to supply Government House (Fort Thornton) with water. The tower is now part of the Parliament Buildings.

The British built three Martello towers in South Africa, one at Simon’s Town naval base, one at Cape Town, and one at Fort Beauford. The Cape Town tower was demolished over a century ago, but that at Simon’s Town, built in 1795, lays claim to being the oldest Martello in the world. That is arguable, because it is not a ‘classic’ Martello.

A Martello tower on St Helena, Napoleon’s last home in exile, was incorporated into High Knoll fort.

One Martello was built in Sri Lanka, at Hambantota on the south coast. There is purely circumstantial evidence that it repelled a French attack. British engineers commenced work on three towers to protect Trincomalee but never completed them.

The last Martello tower built in the British Empire is part of the larger Fort Denison, built on small Pinchgut Island in Sydney Harbour, New South Wales. Fortification of the island began in 1841 following a night-time incursion into the harbour by two American warships, but were not completed. Construction resumed in 1855 to provide Sydney with protection against the threat of a naval attack by the Russian navy during the Crimean War. It was not completed until 1857, well after the war had ended.

The tidal wave which followed the volcanic explosion of Krakatoa in 1883 damaged the Martello tower of Menara that the Dutch East India Company had built in 1850 on Bidadari Island (Pulau Bidadari), a former leper colony, as part of a set of fortifications that protected the approaches to Batavia.

The US government built several Martello towers along the eastern seaboard, two at Key West, and others protecting the harbours of Portsmouth, New Hampshire; Charleston, South Carolina; and New York. Two more sprang up at Tybee Island, Georgia, and Bayou Dupre, Louisiana.

The Americans copied the design from the Canadian Martellos, but there were significant differences. Those on Key West were square instead of round and had thin walls with long gun loops. They were encircled by a curtain wall of heavy guns, making them, effectively, keeps instead of stand-alone towers. The Tybee Island tower built around 1815 was made mainly of wood rather than stone, with gun loops cut into the garrison deck.

Like most Martellos across the world, they were never directly tested in battle.

Han Dynasty Walls

The Han were an ethnically Chinese dynasty founded by a soldier who had nothing to lose as he faced execution for late arrival on parade. It was safer to rebel and bring down what was left of the Qin. The Han first emperor Gaodi (206–195 BC) inherited an exhausted state, barely able to defend itself, let alone launch into giant public works. Gaodi was reported to have ordered the restoration of the great wall but it is difficult to believe that his plan was very effective, if anything happened at all.

The early Han tried several strategies to deal with the resurgent Xiongnu who had returned to the Ordos Desert, from which Qin Shi Huang’s general Meng Tian had expelled them. The Xiongnu then occupied the Gansu Corridor. In 200 BC Gaodi led an attack on them but was ambushed by the Xiongnu leader Maodun, and suffered a complete defeat at Pincheng. There was little then that Gaodi and the early Han emperors could do except pay tribute – sometimes in the form of sending over reluctant princesses – a policy called heqin. Heqin offended Confucian moralists who were disturbed by the improper organisation of the universe, where China should be ‘first under heaven’. But responsibility for the failure of heqin did not lie only with Chinese moralists; it also eventually proved ineffective because the Xiongnu leadership were unable to control all the members of their fissiparous confederation, and this state of affairs resulted in freelance raiding.

Anyway, the early Han had both the qualities of parsimoniousness and durability sufficient to survive sixty years, long enough to replenish the treasuries and turn the tables on the Xiongnu. The later Han sought the wondrous horses of the Ferghana Valley and extended Imperial China along the Gansu Corridor. Emperor Wu (141–87 BC) started to expand the Empire and sent envoys to Central Asia to find allies against the Xiongnu. He recovered the Ordos and conquered the Gansu Corridor.

There was then a renewed campaign of wall-building to secure land retaken from the Xiongnu and to protect the Gansu Corridor to the west. It was during the Han period that China’s walls reached their greatest extent. These walls had a number of functions which were not purely defensive. They protected land that had been retaken from the Xiongnu, so they were partly aggressive in that they supported expansion. They protected the transport route that linked China to Central Asia so the Han could strike at the Western Xiongnu’s power base far from China. They defended the major trade route to Central Asia. Also, much of the eastern part of the Gansu Corridor is loess and there are oases along the route. Thus, one clear function of the walls was also to protect valued irrigated land.

The Western Han collapsed in AD 24 in a civil war which saw the destruction of the capital Xuan. The Han survived from 25 to 220 at the new capital Luoyang and the dynasty become known as the Eastern Han. Defence of Luoyang against the Xiongnu required the further construction of walls in the east.

The key issue is what had fundamentally changed, so that the Han, who had lived without walls for over a century, suddenly went in for a near orgy of wall-building? Under their reign China saw the greatest length of walls constructed – over 10,000 kilometres – at any time in its history. The reason was not initially based on a defensive mentality: rather it was to consolidate expansion at the expense of the Xiongnu, and to guarantee trade routes that extended to Central Asia. Of course, it was important to have the money to fund aggression and growth – and one reason why there was little wall building in the first half of the dynasty’s life was that the Han were refilling their coffers. Furthermore, the heqin policy worked – even if it was deemed rather shameful to Chinese sensibilities.

The Xiongnu also played their part in creating the conditions for renewed wall building. As time passed their increasing claims for brides, bribes and hospitality became less acceptable to an Empire that was recovering the power to reject such claims. Also, the inevitably loose nature of the Confederation meant that, even if agreements held at the centre, it was impossible to control outer tribes who might indulge in some independent raiding. While the Han may have been much more outward looking than many other dynasties, accommodation with barbarian nomads was still seen as demeaning – especially in the face of increasingly arrogant demands and the growing capability to wall them out.

But the process of defeating one set of nomads created its own dialectic through the emergence of another. The effective elimination of the Rong and Di by the three northern states of Yan, Zhao and Qin had provoked the emergence of the Xiongnu. The Han defeat of the Xiongnu by AD 89 meant that the way was clear for another nomad Confederation, the Xianbei, to slip into the breach. Anyway, the Han, like most dynasties, ultimately lost vitality, and the Empire broke up into ethnically Chinese not nomad-ruled states. By the early third century the Ordos was abandoned and the Han collapsed.

The Kugelbunker



The Kugelbunker (Ball or Spherical Bunker) was a late war expedient and may have been derived from the Finnish ‘Ball’ bunker which was based on the idea of an American naval officer. The Kugelbunker had a diameter of 2.13m and could sleep four men, but a smaller version was also more widely used.

One of the final developments in German bunker design was the creation of the Kugelbunker or spherical bunker. Similar in appearance to the Finnish ball bunker, its construction method was different. Few details are available about it beside a post-war report made by the Joint Intelligence Objectives Agency in 1945, which identified one variant. According to this document, in late 1944 Dr Hubert Rusch of the engineering firm Dyckerhoff and Widmann created most of the designs for these bunkers. The army quickly adopted them and ordered several thousand. Production was to be done at about twenty concrete plants in Germany, but the only production centre identified was a Dornbirn, near Lake Constance, where two to three dozen men in each of five concrete plants built them. The largest factory produced six a day. Since production only began in April 1945, all twenty to thirty Kugelbunkers made at Dornbirn went into positions close to the nearby Swiss border. These bunkers consisted of six segments cast in concrete, the top one of which was different since it included a neckpiece that served as an entrance as well as a fighting position. The other five sections were similar, but side entrances could be chiselled into one or two of them after all the segments were cemented together. When the bunker was assembled, its diameter was about 2.1m and its neck was about 37cm high. The interior included a place for four sleeping slabs, although it would not be practical for all four men to stand up at the same time. The man on duty stood on a platform allowing him to occupy the open neck. The wall thickness was 4cm, but the interior was designed for metal reinforcements although none was actually used. The entire bunker weighed less than 2 tons, which facilitated its transportation because several specimens could be placed on a trailer, hauled to the site, and rolled into their excavated position. Since Dornbirn was near the Swiss border, all the bunkers produced there were installed nearby, which made them part of the fictitious National Redoubt.

Little data exists on other types of Kugelbunkers, but it appears that some plants produced a larger number of smaller ones with a diameter of only 1.7m, but a thickness of 10cmm, which had barely enough room for one or two men. Some of these have been found on the Lower Danube front and in Slovakia as part of the defences of the Southeast Front. The bunker on display in the Vienna museum has the entrance at the bottom, but it seems more likely that it would have it on the side or the top from where the occupant could fight as seen in some photos taken in situ. These were the last bunkers of the Third Reich.

Przemyśl, Austria’s Bulwark in the East



In 1914, Festung Przemyśl was among the strongest and most modern of the Austrian fortresses and played a key role in the defence of Galicia. Austria’s first attempt to fortify the city took place early in the Crimean War even though it remained neutral despite efforts to draw it into the war. The old Turkish threat had receded for decades as Russia had advanced towards the Danube. In 1853, Turkish forces fell back and the Russian forces occupied Walachia and Moldavia in 1854 and started driving into Bulgaria. The Austrian Army took up positions in Transylvania, threatening the Russians, considering the Turks to be a buffer rather than a threat. The Turks finally pushed the Russians back and the Austrians administered a neutral Walachia and Moldavia from 1854 until early 1857. During this period of rising tensions, the Austrians had to protect their relatively open frontier with Russia in Galicia. The process began with the construction of seven-sided artillery earthen positions of the FS type. By 1855, a total of nineteen of these positions were completed before construction was interrupted due to improved relations with Russia. With the passing of years, these works deteriorated and no further effort at fortification was undertaken until 1878 except for the construction of an enceinte in 1873, which dragged out into the 1880s. During the 1877 Russo-Turkish War, a number of octagonal FS type positions were placed several kilometres from the city to create a fortified camp.

In 1877, Russia again went to war with the Turks. Austria had agreed to remain neutral in exchange for being allowed to move into Bosnia-Herzegovina. At the end of the war, the Turks lost control over additional territory in the Balkans and Austria-Hungary formalized its right to occupy Bosnia-Herzegovina. Even though the Imperial Fortifications Committee had decided to fortify Przemyśl in 1868, nothing was built there until Russia threatened Galicia once more and Vienna authorized new construction. Construction of an enceinte began in 1873. Building began on nine detached earthen forts in June 1878 and most neared completion by the end of September 1878. After 1880 the army began new work that included rebuilding some of these forts as permanent structures. If a new fort built nearby replaced the old one, the old one became an earthen artillery position. These 1878 forts were on the southern fronts (southeast and southwest front) and on the northern fronts – Fort X ‘Orzechowce’ in the northwest and XIII ‘San Rideau’ in the northeast. Wooden palisades in the moat functioned as obstacles rather than Carnot Walls.

Fort VIII ‘Łętownia’ begun in 1881 was the first permanent artillery fort at Przemyśl. It was built on an old seven-sided FS-type fort from 1854 located on the western front. ‘Łętownia’ was a single-rampart artillery fort with five sides, two caponiers covering the two side ditches, and a double caponier covering the two frontal ditches. A brick wall without embrasures, not a Carnot Wall, stood at the base of the scarp to serve only as an obstacle. Late in the 1880s, Carnot Walls were no longer built. The earth-covered concrete caserne was part of the gorge wall, which also included embrasures for defensive weapons. The central courtyard was occupied by a large shelter built like the barracks and used as the main magazine. The six traverses included munitions and troop shelters adjacent to the ramps that led to the artillery positions between them. There were 120mm M-61 cannons in the open position facing the front, and 90mm M-75 guns on the flanks. Access to the caponiers was by posterns that went through the rampart. Not all of the forts followed this precise design, so each was unique in its own way.

In 1882, construction began on additional single-rampart forts, but it was not completed until 1886. The additional forts included V ‘Grochowce’, VII ‘Pralkowce’, and XII ‘Werner’. They were similar to Fort VIII, but larger. Fort V and VII were on the southwest front and XII on the northeast front. Exact details on all the forts are still being researched since some of their components were very likely incomplete, but the destruction of the forts during the war makes it difficult to know for sure.

The army planned to build several forts into the enceinte, but lack of funding prevented the realization of the project until about 1887. The forts that were eventually built were of a temporary nature, intended to deter enemy cavalry raids, and most had no caserne. During the 1880s, the wooden barracks of the forts and those of the enceinte were covered with a concrete layer. However, these positions were removed during the next decade. In 1887, the shelters used in the forts and other positions were of the Wellbach style in which curved iron sheets formed the interior leaving two open ends like a tunnel that were closed with brick walls. The larger forts included the three double-rampart forts of X ‘Orzechowce’, XI ‘Duńkowiczki’ and XIV ‘Hurko’, and the unusual Fort I ‘Salis-Soglio’ of a non-standard design created by its namesake. This large and unique fort was never outfitted with armoured turrets and had only open positions for its artillery on the ramparts. It had been intended to be the first Panzerwerk in the fortress and was to mount two large Grüson turrets with 120mm guns. However, the armour was cancelled because the War Ministry deemed it too expensive. Most of the forts built in the 1880s became Einheitsforts and became Panzerwerke when armour was added to them in the 1890s. After 1880, concrete roofs were built on all new, renovated, or rebuilt forts.

General von Brunner, Salis-Soglio’s director of fortifications at Przemyśl between 1889 and 1893, was responsible for the construction of the northwest part of the girdle of detached forts between 1892 and 1894. Most of his forts included casernes for about 300 men and some comprised traditors (flanking casemates). The gun batteries consisted of an 80mm gun with a Minimalschartenlafetten (a minimal embrasure mount) that allowed for minimum exposure of the gun, which was placed behind an armoured shield. Thus, Brunner’s Fort IX ‘Brunner’ and XIII ‘San Rideau’, built between 1892 and 1896, became the first Panzerwerke at Przemyśl. Other Panzerwerke with armoured turrets and observation cloches began to appear after these two forts. Fort XIII ‘San Rideau’ was a good example. It included a line of turrets above the caserne with an observation cloche on each end and an artillery observation cloche in the centre with three turrets for 150mm howitzers and mortars on each side of it. The ditch was protected by counterscarp casemates and there was a caponier attached to the caserne in the gorge.

On the rampart, the Panzerwerke generally included a rifle gallery with embrasures for rifles that were covered with armoured plates and sandbags between embrasures for added protection. This gallery was not a covered position although wooden planks were added for overhead protection during combat. In addition to artillery, these forts included machine-gun positions. By the turn of the century, the Austrians began adding iron fences in the gorge and in the ditch in front of caponiers and counterscarp casemates for additional protection. However, at Przemyśl, unlike some forts on the Italian Front, the fences did not cover the entire ditch where once wooden palisades had been used.

In the 1890s, infantry positions called Nahkampfort (close defence forts) were built to fill gaps between the artillery forts. In addition, nine small Panzerwerke armed with either two or four 80mm gun turrets were added. They included I/1 Łysiczka, I/2 Byków, I/5 Popowice, and I/6 Dziewieczyce in the southeast in front of Fort I. The entire complex formed the Siedlisko Group. Forts XIa and I-2, built late in the decade, are laid out according to Brunner’s designs with semicircular shape and a ‘V’-shaped ditch instead of a flat bottom. The gorge was covered by a semicircular earthwork, a centrally located two-level caserne with a combat position and a line of four 80mm gun turrets on its roof, and an observation cloche behind them. The rifle gallery was behind the turrets. Fort I-5 was similar but it had two gun turrets on each side of the block.

The larger Panzerwerke built between 1892 and 1900 included Fort IV ‘Optyń’ with four 80mm gun turrets, Fort IX ‘Brunner’, and Fort XIII ‘San Rideau’ each with three 150mm howitzers and three (four at Fort IX) 150mm mortar turrets. The double-rampart artillery forts were modernized as Panzerwerke in the last few years of the century when X Orzechowce and XI Duńkowiczki were outfitted with four 80mm gun turrets each. These two forts were the only ones at Przemyśl to receive armoured batteries and traditors. Fort IV, the last of the large forts built in 1897–1900 in the fortress girdle, had a lunette shape and included a large two-level caserne in the gorge and a gorge caponier. The barracks accommodated the 450-man garrison (including about 200 artillerymen, 230 infantrymen, and specialists and officers). A large two-level shelter and hangar for the artillery in the centre of the fort connected to the barracks. There were six positions for 150mm guns on the rampart, between the traverses. The fort had traditors – each mounting four 120mm guns – on each side of the rampart. They were located behind the rampart and peeked out of a large embrasure cut into the wall. On the two front corners of the rampart, adjacent to the traditors, there were two turrets with 80mm guns. An observation cloche stood between the turrets and the traditors. The counterscarp casemates at the two front corners mounted machine guns. Additional smaller forts were added after Fort IV.

The roofs of buildings in the last generation of forts were made of concrete poured over steel ‘I’ beams that resulted in flat ceilings rather than the arched ones made with bricks or concrete only. The infantry positions on the ramparts also included two to four positions for light field guns. In the 1900s, the army converted some of the old FS positions and older forts such as at Fort III and VI into infantry positions, some artillery emplacements. In some cases, the ramparts were strengthened with the addition of stone. In 1910, specific instructions were issued to create field fortifications to fill the gaps in the ring and link them with a continuous line of trenches. However, only two infantry strongpoints were built before the war and several after the war began. These strongpoints included wooden lined trenches and shelters designed to house forces ranging from half an infantry company to two companies.

Construction on the existing forts continued into the early twentieth century. Forts II, III, and VI were among the last to be rebuilt and modernized. By 1914, the Przemyśl was, if not the largest, at least one of the most important and modern fortresses in the Empire.

Constantinople Prepares for Siege



Winters on the Bosphorus can be surprisingly severe, as the Arabs had discovered during the siege of 717. The site of the city, jutting out into the straits, leaves it exposed to fierce squalls hurtling down from the Black Sea on the north wind. A particularly dank and sub-zero cold penetrates to the marrow of the bones; weeks of cheerless rain can churn the streets into mud and prompt flash floods down the steep lanes; sudden snowstorms arise as if from nowhere to obliterate the Asian shore half a mile away then vanish as quickly as they have come; there are long still days of muffling fog when an eerie silence seems to hold the city in an iron grasp, choking the clappers in church bells and deadening the sound of hooves in the public squares, as if the horses were shod in boots of felt. The winter of 1452–53 seems to have afflicted the citizens with particularly desolate and unstable weather. People observed ‘unusual and strange earthquakes and shakings of the earth, and from the heavens thunder and lightning and awful thunderbolts and flashing in the sky, mighty winds, floods, pelting rain and torrential downpours’. It did not improve the overall mood. No flotillas of Christian ships came to fulfil the promises of union. The city gates remained firmly closed and the supply of food from the Black Sea dried up under the sultan’s throttle. The common people spent their days listening to the words of their Orthodox priests, drinking unwatered wine in the taverns and praying to the icon of the Virgin to protect the city, as it had in the Arab sieges. A hysterical concern for the purity of their souls seized the people, doubtless influenced by the fulminations of Gennadios. It was considered sinful to have attended a liturgy celebrated by a unionist or to have received communion from a priest who was present at the service of union, even if he were simply a bystander to the rites. Constantine was jeered as he rode in the streets.

Despite this unpromising atmosphere, the emperor made what plans he could for the city’s defence. He dispatched envoys to buy food from the Aegean islands and beyond: ‘wheat, wine, olive oil, dried figs, chick peas, barley and other pulses’. Work was put in hand to repair neglected sections of the defences – both the land and sea walls. There was a shortage of good stone and no possibility of obtaining more from quarries outside the city. Materials were scrounged from ruined buildings and abandoned churches; even old tombstones were pressed into service. The ditch was cleared out in front of the land wall and it appears that despite their reservations, Constantine was successful in persuading the populace to participate in this work. Money was raised by public collection from individuals and from the churches and monasteries to pay for food and arms. All the available weapons in the city – of which there were far too few – were called in and redistributed. Armed garrisons were dispatched to the few fortified strongholds still held by Byzantium beyond its own walls: at Selymbria and Epibatos on the north shore of the Marmara, Therapia on the Bosphorus beyond the Throat Cutter, and to the largest of the Princes’ Islands. In a final gesture of impotent defiance, Constantine sent galleys to raid Ottoman coastal villages on the Sea of Marmara. Captives were taken and sold in the city as slaves. ‘And from this the Turks were roused to great anger against the Greeks, and swore that they would bring misfortune on them.’

The only other bright spot for Constantine during this period was the arrival of a straggle of Italian ships that he was able to persuade – or forcibly detain – to take part in the city’s defence. On 2 December a large Venetian transport galley from Kaffa on the Black Sea, under the command of one Giacomo Coco, managed to trick its way past the guns at the Throat Cutter by pretending that it had already paid its customs dues further upstream. As it approached the castle the men on board began to salute the Ottoman gunners ‘as friends, greeting them and sounding the trumpets and making cheerful sounds. And by the third salute that our men made, they had got away from the castle, and the water took them on towards Constantinople.’ At the same time news of the true state of affairs had reached the Venetians and Genoese from their representatives in the city and the Republics stirred themselves into tardy activity. After the sinking of Rizzo’s ship, the Venetian Senate ordered its Vice-captain of the Gulf, Gabriel Trevisano, to Constantinople to accompany its merchant convoys back from the Black Sea. Among the Venetians who came at this time was one Nicolo Barbaro, a ship’s doctor, who was to write the most lucid diary of the months ahead.

Within the Venetian colony in the city, concern was growing. The Venetian bailey, Minotto, an enterprising and resolute man, was desperate to keep three great merchant galleys and Trevisano’s two light galleys for the defence of the city. At a meeting with the emperor, Trevisano and the other captains on 14 December he begged them to stay ‘firstly for the love of God, then for the honour of Christianity and the honour of our Signoria of Venice’. After lengthy negotiations the ships’ masters, to their credit, agreed to remain, though not without considerable wrangling over whether they could have their cargo on board or should keep it in the city as surety of their good faith. Constantine was suspicious that once the cargo was loaded, the masters would depart; it was only after swearing to the emperor personally that they were allowed to load their bales of silk, copper, wax and other stuffs. Constantine’s fears were not unfounded: on the night of 26 February one of the Venetian ships and six from the city of Candia on Crete slipped their anchors and fled before a stiff north-easterly wind. ‘With these ships there escaped many persons of substance, about 700 in all, and these ships got safely away to Tenedos, without being captured by the Turkish armada.’

This dispiriting event was offset by one other positive contribution. The appeals of the Genoese podesta at Galata had elicited a concrete offer of help. On about 26 January two large galleons arrived loaded ‘with many excellent devices and machines for war, and outstanding soldiers, who were both brave and confident’. The spectacle of these ships entering the imperial harbour with ‘four hundred men in full armour’ on deck made an immediate impression on both the populace and the emperor. Their leader was a professional soldier connected to one of the great families of the republic, Giovanni Giustiniani Longo, a highly experienced commander who had prepared this expedition at his own initiative and cost. He brought 700 well-armed men in all, 400 recruited from Genoa, another 300 from Rhodes and the Genoese island of Chios, the power base of the Giustiniani family. Constantine was quick to realize the value of this man and offered him the island of Lemnos if the Ottoman menace should be repulsed. Giustiniani was to play a fateful role in the defence of the city in the weeks ahead. A straggle of other soldiers came. Three Genoese brothers, Antonio, Paolo and Troilo Bocchiardo brought a small band of men. The Catalans supplied a contingent and a Castilian nobleman, Don Francisco of Toledo, answered the call. Otherwise the appeal to Christendom had brought nothing but disharmony. A sense of betrayal ran through the city. ‘We had received as much aid from Rome as had been sent to us by the sultan of Cairo,’ George Sphrantzes recalled bitterly.

With the arrival of the Genoese contingent the preparations for a siege were carried forward with greater urgency. Giustiniani, who was ‘an expert in the art of wall fighting’, appraised the city’s defences with a cool eye and took appropriate measures. Under his direction, during February and March they ‘dredged the fosse and repaired and built up the walls, restoring the battlements, refortifying inner and outer towers and strengthening the whole wall – both the landward and seaward sectors’.

Despite their dilapidated condition, the city still possessed formidable fortifications. Among all the many explanations for the longevity of Byzantium, the impregnable defences of its capital city remain a cardinal factor. No city in the world owed as much to its site as Constantinople. Of the twelve miles of its perimeter, eight were ringed by sea. On the south side, the city was fringed by the Sea of Marmara, whose swift currents and unexpected storms made any sea-borne landing a risky undertaking. In a thousand years no aggressor ever seriously attempted an attack at this point. The seashore was guarded by a single unbroken wall at least fifty feet above the shoreline interspersed with a chain of 188 towers and a number of small defended harbours. The threat to this wall came not from ships but from the ceaseless action of the waves undermining its foundations. At times nature was more brutal still: in the bitter winter of 764 the sea walls were crushed by ice floes that rode up over the parapets. The whole length of the Marmara wall was studded with marble inscriptions commemorating the repairs of successive emperors. The sea ran strongly round this shoreline as far as the tip of the Acropolis point, before turning north into the calmer waters of the Golden Horn. The Horn itself provided an excellent sheltered anchorage for the imperial fleet; 110 towers commanded a single wall along this stretch with numerous water gates and two substantial harbours, but the defences were always considered vulnerable. It was here that the Venetians had driven their ships up on the foreshore during the Fourth Crusade, overtopping the ramparts and storming the city. In order to block the mouth of the Horn in times of war, the defenders had been in the habit, since the Arab siege of 717, of drawing a boom across the entrance of the Horn. This took the form of a 300-yard chain, consisting of massive cast-iron links each twenty inches long that were supported on sturdy wooden floats. With the good will of the Genoese, the chain could then be secured to a tower on the sea wall of Galata on the far side. During the winter months both chain and floats were prepared against the possibility of a naval attack.

The base of the triangle of the city’s site on the westward side was protected by the four-mile land wall, the so-called wall of Theodosius, which ran across the grain of the land from the Sea of Marmara to the Golden Horn and sealed off Constantinople from any conventional land-borne assault. Many of the most significant events in the history of the city had been played out along this extraordinary structure. It almost matched the city itself in longevity, and projected a sense of legendary immutability within the Mediterranean world. For many approaching Constantinople across the flat Thracian plains as a trader or pilgrim, an ambassador from a Balkan court or a plundering army with pretensions to conquest, the first sight of Constantinople at its apogee would be the ominous prospect of the land walls riding the gentle undulations of the landscape from horizon to horizon in a regular unbroken succession of ramparts and towers. In the sunlight the limestone walls create a facade of brilliant white, banded with horizontal running seams of ruby red Roman brick, and arrow slits similarly arched; the towers – square, hexagonal, octagonal, occasionally circular – are so close together that, as one crusader put it, ‘a seven-year-old boy could toss an apple from one turret to the next’. They rise up in successive tiers to the summit of the inner wall where the eagle banners of the emperor flutter proudly in the wind. At intervals the eye can pick out the darkness of a heavily guarded entrance to the city through which men and animals vanish in times of peace, and at the western end, close to the Sea of Marmara, a gateway panelled with flat plates of gold and decorated with statues of marble and bronze shines in the sun. This is the Golden Gate, the great ceremonial archway flanked by two massive towers of polished marble through which, in the heyday of Byzantium, emperors returned in triumph with the visible tokens of their victories: conquered kings walking in chains, recaptured sacred relics, elephants, outlandishly dressed barbarian slaves, carts piled high with booty and the whole might of the imperial army. By 1453 the gold and many of the decorations had gone but the structure was still an impressive monument to Roman glory.

The man responsible for the land wall, built to define the mature limits of the city, was not the boy Emperor Theodosius after whom it is named, but a leading statesman of the early fifth century, Anthemius, ‘one of the wisest men of the age’, for whose far-sightedness the city would owe a limitless debt of gratitude. The first line of the walls built in 413 deterred Attila the Hun, ‘the scourge of God’, from making an attack on the city in 447. When it collapsed under a severe earthquake the same year with Attila ravaging Thrace not far away, the whole population responded to the crisis. Sixteen thousand citizens totally rebuilt the wall in an astonishing two months, not just restoring Athemius’s original structure, but adding an outer wall with a further string of interspaced towers, a protecting breastwork and a brick-lined moat – the fosse – to create a formidable barrier of extraordinary complexity. The city was now protected on this side by a chain of 192 towers in a defensive system that comprised five separate zones, 200 feet wide and 100 feet high from the bed of the moat to the top of the tower. The achievement was recorded with a suitably boastful inscription: ‘In less than two months, Constantine triumphantly set up these strong walls. Scarcely could Pallas have built so quickly so strong a citadel.’

In its mature form, the Theodosian wall summarized all the accumulated wisdom of Graeco-Roman military engineering about defending a city before the age of gunpowder. The heart of the system remained the inner wall constructed by Anthemius: a core of concrete faced on both sides by limestone blocks quarried nearby, with brick courses inserted to bind the structure more firmly. Its fighting ramparts were protected by battlements and reached by flights of steps. In line with Roman practice, the towers were not bound to the walls, ensuring that the two structures could each settle at their own rate without breaking apart. The towers themselves rose to a height of sixty feet and consisted of two chambers with a flat roof on which engines to hurl rocks and Greek fire could be placed. Here the sentinels scanned the horizon unceasingly, keeping themselves awake at night by calling out to each other down the line. The inner wall was forty feet high; the outer one was lower, about twenty-seven feet high, and had correspondingly lower towers that interspaced those on the inner wall. The two walls were separated by a terrace sixty feet wide, where the troops defending the outer wall massed, ready to engage the enemy at close quarters. Below the outer wall another terrace sixty feet wide provided a clear killing field for any aggressor who made it over the moat. The brick-lined moat itself was another sixty-feet-wide obstacle, surmounted by a wall on the inner side; it remains unclear whether it was in parts flooded in 1453 or simply comprised a dry ditch. The depth and complexity of the system, the stoutness of its walls and the height from which it commanded its field of fire rendered the Theodosian wall virtually impregnable to an army equipped with the conventional resources of siege warfare in the Middle Ages.

Along its length the land wall was pierced by a succession of gates. Some gave access to the surrounding countryside via bridges over the moat, which would be destroyed in the run-up to a siege; others, the military gates, allowed connection between the different layers of the walls and were used to move troops about within the system. The wall also contained a number of posterns – small subsidiary doorways – but the Byzantines were always aware of the danger these sally ports posed for the security of their city and managed them rigorously. In general the two sets of gates alternated along the length of the wall, with the military gates being referred to by number while the public gates were named. There was the Gate of the Spring, named after a holy spring outside the city, the Gate of the Wooden Circus, the Gate of the Military Boot Makers, the Gate of the Silver Lake. Some spawned multiple names as associations were forgotten and new ones created. The Third Military Gate was also referred to as the Gate of the Reds, after a circus faction in the early city, while the Gate of Charisius, a leader of the blue faction, was also called the Cemetery Gate. And into the structure were built some remarkable monuments that expressed the contradictions of Byzantium. Towards the Golden Horn the imperial palace of Blachernae nestled behind the wall, a building said once to be of such beauty that foreign visitors could find no words to describe it; adjoining it, the dank and dismal prison of Anemas, a dungeon of sinister reputation, scene of some of the most ghastly moments in Byzantine history. Here John V blinded both his son and his three-year-old grandson, and from here one of Byzantium’s most notorious emperors, Andronikos the Terrible, already horribly mutilated, was led out on a mangy camel amongst taunting crowds to the Hippodrome, where he was strung upside down between two columns and mockingly slaughtered.

The continuous life of the wall was so long that a deep accretion of history, myth and half-forgotten association attached to the various sectors. There was hardly a place that had not witnessed some dramatic moment in the city’s history – scenes of terrible treachery, miraculous deliverance and death. Through the Golden Gate Heraclius brought the True Cross in 628; the Gate of the Spring saw the stoning of the unpopular Emperor Nicephorus Phocas by an enraged mob in 967 and the restoration of the Orthodox emperors after Latin rule in 1261 when the gate was opened from within by sympathizers. The dying Emperor Theodosius II was carried through the Fifth Military Gate in 450 following a fall from his horse in the valley outside, while the Gate of the Wooden Circus was blocked up in the twelfth century after a prophecy that the Emperor Frederick Barbarossa would use it to capture the city.

Next to St Sophia itself no structure expressed the psychic life of the city’s people as powerfully as the walls. If the church was their vision of heaven, the wall was their shield against the battering of hostile forces, under the personal protection of the Virgin herself. During sieges the constant prayers and the procession of her sacred relics along the ramparts were considered by the faithful to be generally more crucial than mere military preparations. A powerful spiritual forcefield surrounded such actions. Her robe, housed at the nearby church at Blachernae, was accorded more credit for seeing off the Avars in 626 and the Russians in 860 than military engineering. People saw visions of guardian angels on the ramparts and emperors inserted marble crosses and prayers into the outward facing walls. Near the centre point of the wall there is a simple talisman that expresses Constantinople’s deepest fear. It says: ‘O Christ God, preserve your city undisturbed and free from war. Conquer the fury of the enemies.’

At the same time, the practical maintenance of the walls was the one essential public work for the city, in which every citizen was required to help, without exemption. Whatever the state of the Byzantine economy money was always found to patch up the wall. It was sufficiently important to have its own special officials under the overall authority of the impressively named ‘Count of the Walls’. As time and earthquakes shattered towers and crumbled masonry, running repairs were marked by a wealth of commemorative marble inscriptions set into the stonework. They span the centuries from the first reconstruction in 447 to a total renovation of the outer wall in 1433. One of the last dated repairs before the siege expresses the co-operation of divine and human agencies in the maintenance of the city’s shield. It reads: ‘This God-protected gate of the life-giving spring was restored with the co-operation and at the expense of Manuel Bryennius Leontari, in the reign of the most pious sovereigns John and Maria Palaeologi in the month of May 1438.’

Perhaps no defensive structure summarizes the truth of siege warfare in the ancient and medieval world as clearly as the walls of Constantinople. The city lived under siege for almost all its life; its defences reflected the deepest character and history of the place, its mixture of confidence and fatalism, divine inspiration and practical skill, longevity and conservatism. Like the city itself, the walls were always there, and for anyone in the eastern Mediterranean, it was assumed they always would be. The structure of the defences was mature in the fifth century and changed little thereafter; the building techniques were conservative, harking back to practices of the Greeks and Romans. They had no particular reason to evolve because siege warfare itself remained static. The basic techniques and equipment – blockade, mining and escalade, the use of battering rams, catapults, towers, tunnels and ladders – these were largely unchanging for longer than anyone could recall. The advantage always lay with the defender; in the case of Constantinople its coastal position increased that weighting. None of the armies camped before the land walls had ever succeeded in effecting an entry through the multiple defensive layers, while the city always took prudent measures as a matter of state policy to keep its cisterns brimming and its granaries full. The Avars came with an impressive array of stone-throwing machinery but their looping trajectory made them far too puny to breach the walls. The Arabs froze to death in the cold. The Bulgar Khan Krum tried magic – he performed human sacrifices and sprinkled his troops with seawater. Even its enemies came to believe that Constantinople was under divine protection. Only the Byzantines themselves were ever successful in taking their own city from the land, and always by treachery: the messy final centuries of civil war produced a handful of instances where gates were flung open at night, usually with inside help.

There were just two places where the land wall could be considered potentially weak. In the central section the ground sloped down a long valley to the Lycus River and then up the other side. As the wall followed the downward slope, its towers no longer commanded the high ground and were effectively below the level occupied by a besieging army on the hill beyond. Furthermore the river itself, which was ducted into the city through a culvert, made it impossible to dig a deep moat at this point. Nearly all besieging armies had identified this area as vulnerable, and though none had succeeded, it provided attackers with a vestige of hope. A second anomaly in the defences existed at the northern end. The regular procession of the triple wall was suddenly interrupted as it approached the Golden Horn. The line took an abrupt right-angle turn outwards to include an extra bulge of land; for 400 yards, until it reached the water, the wall became a patchwork structure of different-shaped bastions and sectors, which, though stoutly built on a rocky outcrop, was largely only one line deep and for much of its length unmoated. This was a later addition undertaken to include the sacred shrine of the Virgin at Blachernae. Originally the church had been outside the walls. With a typical Byzantine logic it had been held initially that the protection of the Virgin was sufficient to safeguard the church. After the Avars nearly burned it in 626 – the shrine was saved by the Virgin herself – the line of the wall was altered to include the church, and the palace of Blachernae was also built in this small bight of land. Both these perceived weak spots had been keenly appraised by Mehmet when he reconnoitred in the summer of 1452. The right-angle turn where the two walls joined was to receive particular attention.

As they patched up their walls under Giustiniani’s direction and paraded the sacred icons on the ramparts, the people of the city could be pardoned for expressing confidence in their protective powers. Immutable, forbidding and indestructible, they had proved time and again that a small force could keep a huge army at bay until its willpower collapsed under the logistical burden of siege, or dysentery or the disaffection of the men. If the walls were decayed in places, they were still basically sound. Brocquière found even the vulnerable right angle to be protected by ‘a good and high wall’ when he came in the 1430s. The defenders however were unaware that they were preparing for conflict on the cusp of a technological revolution that would profoundly change the rules of siege warfare.