Citadels of Christendom
The Fortress, Krac des Moab (Kerak)
It has been seven hundred years since the last Crusader departed the Middle East for Europe. Behind them, scattered from southern Turkey to the borders of Egypt, they left behind a series of stone fortifications still known collectively as Crusader castles. The remnants of these once formidable obstacles to Saracen power now stand forlorn guard at harbor entrances and along windswept ridges. For two hundred years they were the key to European control of Jerusalem and the rest of what they called the Holy Land. All that remains of some are the foundations and a few courses of stone blocks. Others like Krac des Chevaliers in Syria and Jordan's Kerak are nearly intact. But no matter what the state these remnants of the Franks' occupation are among the finest examples of military architecture ever raised.
When, in a fiery speech to the Council of Clermont in 1095, Pope Urban II declared Deus Vult! (God wills it!) He set in motion the events that inspired Christendom to free the Holy Land from the infidel Turks. The First Crusade (1097-99), driven by religious fervor, passion and aided by an unprepared Turkish defense, swept forward and captured Jerusalem. With Zion now safely in Christian hands the majority of Crusaders returned to Europe. Remaining behind were an odd collection of penitents, military monks, adventurers and free-booters. Forming the Latin Kingdom of Jerusalem and a series of lesser states, they were strangers in a strange land.
From the beginning the Crusaders were heavily outnumbered by the natives. There never seemed to be enough knights to protect the coastal cities and the countryside from the unending Muslim counterattacks. A desperate necessity to augment their small numbers meant that in the design and construction of these fortifications the Crusaders were trying to make stones do the work of men. The shortage of manpower made the castles of the Levant more important to the Crusader military than those of Europe were to their counterparts. In essence the Crusader and his castle were symbiotes, each dependent on the other for survival.
Before addressing the military aspects of Crusader castles two basic points need to be kept in mind.
First, the apparent haphazard distribution of castles is exactly that. There was no central plan to the placement of castles nor are they part of an organized defensive system. There simply was no leader with the power to enforce a plan even if one had been developed. The King of Jerusalem was nominally head of vast territories but he had to contend with major and minor nobles, some under his control, some not. In the more remote areas of the kingdom, a head strong noble might simply ignore him arranging things as best suited him. In addition the fortifications of the two great military monastic orders -- the Knights Templar and the Knights Hospitaller -- were placed where there leadership deemed it best for the good of the order.
Principles of Early Design: The Primacy of the Defense
Fortification design was, and still is, part of the ongoing dialectic between the defense and the offense. Faced with formidable defenses the attacker tries to find ways to overcome them. Those on the defense know the strengths of the attacker and try to devise ways to neutralize it. Hence the architecture of castles (or any fortification) is in actuality a history of siege warfare. From rudimentary beginnings the castles of the Holy Land became sophisticated killing grounds -- weapons systems designed to enable a few defenders to hold out for as long as possible, while selling their lives at as dear a price as they could.
Scholars debate the sources of Crusader castle architecture. Some argue that since the European invaders had only primitive fortresses in the West subsequent Levantine designs were based on Eastern, particularly Muslim and Byzantine, influences. Some, such as T. E. Lawrence, have seen in the designs of the Crusaders elements that were eventually taken to Europe and incorporated there. Other scholars argue, just as vehemently, that though Crusader design incorporated Muslim and Byzantine elements it contained just as many unique elements of its own. Bolstering their argument is the presence of the ruins of literally hundreds of Roman forts throughout Europe. These scholars also reject the idea of technology transference, pointing out that European castle architecture was slow to change in response to the lessons the Crusaders had learned. In reality it seems most likely that the Crusaders used Roman, Byzantine, Muslim, European and their own experiences to develop the designs that best suited their needs.
At the time of the First Crusade European castles were primitive, their design rudimentary. The area occupied by the garrison was surrounded by a ditch, earth bank and timber stockade. Within this bailey a strong point was built by digging a circular ditch and throwing the soil towards the center. The summit of this mound or motte was raised a considerable height above the bottom of the ditch and could be defended by a stockade or even a wooden tower. Thousands of these mottes survive in Europe and the defenses they once held are illustrated in the Bayeux Tapestry. Their real strength lies in the inaccessibility provided by the height of the motte and the steepness of the sides. By the start of the 11th Century the use of stone for the construction of the motte was becoming common, and the motte was crowned by a keep.
The normally took the form of a rectangular tower. Defenses were wholly passive. The besieged and besiegers could not come to grips and the besieged usually sat within their massive defenses waiting for the relatively ill equipped besiegers to withdraw. All of these early principles formed the design basis for the earliest Crusader castles. Except for the coastal areas, castles were usually placed on the highest ground available. This was partly for inaccessibility, partly to force besiegers to labor through an inhospitable environment and partly to allow signaling to the next fortress. Once the site was selected, construction of the castle began. Crusaders always used stone rather than wood in their construction. Partly this was because timber was scarce in the Middle East while stone and stone cutters were plentiful. But primarily the use of stone limited the effect of fire, an important factor since the Muslims were experts in the use of incendiary devices.
Early castles usually had a central tower keep which formed the citadel of the fortress. This keep had a plain exterior with little or no buttressing or other projecting features. The defensive strength of the keep rested with the thickness of its walls in which openings were kept to a minimum. At Saone, for example, the wall of the keep is 15 feet thick at the point of most likely attack. At ground floor level there are only two openings through which it was possible to watch or shoot at an approaching enemy. Within the keep flights of stone steps provided access to the different levels and the roof. But a castle is more than a keep. There needed to be a place for people. A curtain wall of stone block was built to enclose a large area consisting of courtyards where daily life could be conducted and the population could fall back to if the need arose.
Second, while the principal function of a castle is defense, castles played more than a simple defensive role. Fortified building were built and occupied because of the many uses to which they could be put. Castles were the residence of a magnate and his household. They were also the center from which he administered his estates. The great enclosed spaces were markets, gathering places and souks (trading emporia). They echoed more to the footsteps of ordinary people doing everyday things than to the iron clang of armored men rushing to the defense. Any castle of the religious military orders was also a monastery, and the major ones of the Templars and Hospitallers also served as the regional headquarters of what were essentially great corporations.
Castles also played a role in the social and economic development of the Crusader states. Their garrisons were a means to bring security to a region. Castles meant that new settlers could be brought to areas which had previously been uncultivated because they were too exposed to enemy raiding. The 13th Century historian William of Tyre emphasized that Baldwin I's construction of Montreal (al-Shaubak) in 1115 was due to his desire to colonize Oultre Jordain (southern Jordan). Similarly the castle used by Raymond of Gilles as the base for his long term siege of medieval Tripoli in Lebanon became the center of the modern town by the same name.
Balancing the military and other aspects of a castle required careful design as exemplified by the entrance to Krac de Chevaliers, a 13th Century fortress of the Knights Hospitaller. The entry path rises gently from the outer gate to the inner one in a series of broad shallow steps carefully designed for easy use by both men and horses. Along this path the architects added difficulties for an attacker. An enemy fighting his way through the outer gate was slowed by three sudden changes of direction -- a left turn through a right angle followed after a sharp ascent by a hairpin turn. Rapid changes of lighting from dark to bright sunlight and back would obscure his vision. Throughout the approach he was exposed to missiles from overhead, while defenders sallied forth from recesses in the wall. But to the average visitor or tradesmen these same obstacles provided no serious problem and were probably not even noticed.
As time passed the curtain wall and its supporting towers and bastions assumed more importance and designer dispensed with the keep or replaced it with strongly fortified gatehouses
In the early 11th Century the only means of taking a fortification quickly was by a direct assault in which surprise was the critical element. If this failed, and it usually did, then the only option for the attacker was blockade -- surrounding the fortress and attempting to starve it into submission.
Not surprisingly one of the highest priorities for castle architects was the accumulation of large reserves of food and water. Water was far and away the most essential need, both for men and the horses on which any counterattack depended. There were three ways of providing water to a castle: wells, open cisterns and closed cisterns.
Wells were usually fed by an underground spring. At Montreal (al-Shaubak) this required drilling through the top of the mountain to its base and constructing a 365 step narrow staircase to the underground spring passing under the hill. Cisterns, both open and closed, accumulated run off from the occasional rains. The size of these cisterns was often enormous. The ones at Krac de Chevalier could hold a total of 2 million gallons (10 million liters) while those at Kerak were only slightly smaller. Amazingly, in view of the terrain and climate they were in, no Crusader castle ever fell due to lack of water.
Food supplies were equally important and the need just as vast. Margat maintained a five year food supply. Kerak was provisioned for two years. The larger castles, such as Kerak and Saone, were equipped to prepare large amounts of raw grain, both millstones and ovens have been discovered in both. Krac de Chevaliers even has one tower equipped with a windmill.
Other design changes were incorporated into the outer wall of the castle to provide more than a simple passive defense and allow the besieged to harry the besiegers. Curtain walls were reinforced with projecting towers, a Crusader innovation. By placing the tower so it was a few feet out from the wall the tower defenders were able to cover more ground and establish a field of fire against enemy forces attempting to approach the wall with scaling ladders. Bastions, distinguished from towers because they were the same height as the wall, served a similar purpose. Arrow slits were cut into both structures allowing archers a fairly wide field of fire while protecting him against all but the most skilled (or luckiest) enemy bowmen. Early towers and bastions were square or rectangular in shape, a form which was easy to build and did not require skilled stone cutters.
Inner wall were always higher than the outer walls. This allowed those on the inner walls to shoot over the heads of their comrades or fire down on an enemy in the event a portion of the wall was threatened. Often the wall walkways were accessible only by doorways in the flanking towers -- hence even if an attacker did gain a foothold on the curtain wall it did not get him anywhere. The tower doors, while weak by comparison with those at ground level, were easily able to withstand anything the assailant could bring up a scaling ladder.
It was also common practice to smooth the existing sides of any uphill approaches, thus denying the attacker handholds or footholds if he attempted to scale the hill and approach the castle from another direction. Where the castle was built on flat ground, such as in coastal areas, they were given a sloping apron or plinth known as a talus that extended from the outer walls. The talus added strength and height to the walls and prevented the use of scaling ladders since they could not be footed against its angled slope. If placed on more distant level ground the ladders were either too short or resting at such an oblique angle that they were weakened to the point of collapse if several armed men tried to ascend at the same time.
Another advantage of the talus and glacis was that when the defenders dropped stones on to it they splintered and ricocheted with shrapnel effect. To allow the men on the parapets to do this without exposing themselves to arrows, stone verandas were built out from the top of walls with holes in the floors. Called machicolations these "murder holes" became a feature of all castle design from about 1120 onwards.
Internally the castle was a maze and designed for maximum fire on any point within. In addition to the classic "bent entrance" (which the Crusaders were the only ones to use and do not appear in European castles) there were portcullises and drawbridges, dead end corridors and covered walkways that allowed the outer curtain walls to be reinforced or evacuated in safety. The outer, middle and inner wards were usually further subdivided by walls. Against this defensive system all the attacker could do was blockade. But a siege is both resource intensive and time consuming. While waiting for the defenders to starve an attacker had to keep his men and horses fed and watered. Several Muslim accounts record the sheer frustration experienced by the besiegers in trying to reduce strong fortifications by blockade. The besiegers had little if any opportunity to plunder the surrounding area, were subject to fire from the enemy castle and were generally miserable.
Time was also a problem. The normal Muslim campaigning season was spring and summer. By fall the men were anxious to return home. The prospect of a winter in cold and soggy tents was never appealing and the mood of the army was such that a wise commander withdrew it. Also affecting the time factor was the eventual arrival of a relief force bent on breaking the siege. The besiegers could remain in place, as Richard the Lion-hearted did at Acre, but normally withdrew with the arrival of a larger, and fresher army, on the field. Attackers therefore needed to find a way to speed things along. They replaced assault with battery.
During the Crusader siege of Antioch (1097-8) there is no mention of siege weapons. By the time they reached Jerusalem the Crusaders had begun to utilize both the covered battering ram and the siege tower. These were built by Genoese sailors pressed into service because of their skill in working with timber and rope. The covered battering ram was designed to bring down the fortress gate by knocking it down. A large leather awning provided projection from overhead fire. This quaint device, popular in movies, does not appear to ever actually knocked down the entrance of any significant fortress or city.
The siege tower was a ramshackle affair built of timber to the height of the enemy's fortress wall. This was moved forward on rickety wheels and allowed the attackers to gain access to the citadel. Pouring in, the Crusaders would then spread out through the area, open the city gates (allowing entry to the men on the battering ram) and defeat the enemy. In practice siege towers were rarely effective, though at Jerusalem the small size of the defending Turkish garrison allowed the siege tower its brief moment of glory. They were cumbersome, likely to collapse and being made of timber easily destroyed by incendiaries.
Saracen siege engine technology was far ahead of the Crusaders and began to manifest itself in the latter part of the 12th Century. Disdaining both the battering ram and the siege tower the Muslims utilized stone throwing artillery.
It is a mistake to think of stone throwing catapults as Rube Goldberg affairs -- slightly ridiculous mechanisms worked by string and rubber band. It is an even bigger error to think that artillery was only a serious threat to a castle with the coming of gunpowder.
There were three types of stone throwing artillery known to the Franks: the margnel, the petrary and the trebuchet. The first two used torsion or twisted ropes to propel a stone missile while the trebuchet operated on a swing beam. There is no evidence that any but the latter, called a manjaniq in Arabic, was actually used during the Crusader era.
The trebuchet consisted of a long beam placed over a fulcrum with one end projecting much further from the fulcrum than the other. A stone or other item (such as a carcass, corpse or head) was placed in a sling at the end of the long end. The short end was then brought down sharply causing the tip of the long end to rise much faster to a vertical position. The sling at the end of the long end whipped around to release the projectile at great speed. Artillery of this type, probably derived from China, was first described in the Mediterranean in 597 and had been used by Arabic forces during their conquests in the 7th Century. Viking raiders had used trebuchets at the siege of Paris in 885-86. These early devices were easily assembled, portable but not particularly powerful.
In the 12th Century Byzantine engineers made a major design change in the trebuchet. Previously the short end of the fulcrum had been brought down by a gang of men pulling on ropes. They were now replaced by a heavy counterweight, usually a net or box of rocks or sand. The effect was dramatic. The range was increased and the accompanying increase in velocity caused greater damage. The standardized counterweight allowed the trebuchet to maintain a constant aim, repeatedly striking the same area of a wall or a tower.
Its new power was formidable. Normally the ammunition consisted of a specially rounded stone weighing between 100-600 lb. (based on those found at Saone) which it could hurl from 500-700 feet. The new range allowed the artillery safety from opposing archery whose maximum bow shot was about 400-450 feet. The destructive force of the projectile was immense and while it could not take down a wall with a single blow it could, given uninterrupted time, eventually bring down the thickest wall or tower. In addition the psychological impact of continuous bombardment was profound.
But the most feared weapon was the mine. Unless water defenses, bedrock or physical inaccessibility made it impossible there simply was no effective defense against the mining gallery. Early mining operations, such as those conducted by Saladin, were little more than surface sapping. The process became fully developed in the middle of the 13th Century and was the favored method attack methods thereafter. The Saladin era mines started within fifty feet of the target wall or tower. Sappers would usually remove a quantity of earth from the foundations of a corner of a tower. The hole was then filled with bundles of wood, which were then ignited. The fire would crack the mortar in the tower and it would collapse, creating a breach in the wall. While effective, early mining had serious drawbacks. The work was skilled, expensive, dangerous and time consuming. The proximity of early mining operation to the walls meant that they could easily be detected and destroyed. In addition not all castles were subject to assault by this method, those along the coast or protected by a water moat provided obstacles that could not be overcome.
Back to the Drawing Board
The trebuchet and the mine changed the balance from the primacy of the defense to the primacy of the attacker. Castle architects scurried to counter the new attack methods. The biggest immediate concern was the trebuchet. Uninterrupted firing time meant that the walls would eventually come tumbling down as certainly an incoming tide washes away a child's sand castle. The Crusaders saw evidence of this first hand during their siege of Acre in 1191 when their use of trebuchets caused substantial damage to the city walls, considerably thicker than that of any castle. Sitting passively behind the walls while relying on the fortification for defense until the enemy left was no longer possible -- means had to be found to neutralize artillery and counter mines.
The first means was to add trebuchets to the castle's defenses. Tower roofs were flattened and artillery mounted atop them to maximize range and keep the attacker's weapons at a distance. During the 1183 siege of Kerak the defenders were hampered by an inability to respond to Saladin's artillery. When he returned the following year, the defending Crusaders were able to keep his artillery at bay by counterbattery fire of their own. During the siege of Margat in 1285 an artillery duel ensued resulting in the destruction of Muslim artillery. New emphasis was placed on denying the enemy proximity to the walls. Where the castle was on a mountain or hill the walls were brought to the edge of the cliff. At Kerak, where the existing set of walls had left enough flat ground near the three sides of the castle protected on the spur to worry designers, a new curtain wall was constructed and connected to the old walls by vaulted arches. The walls of Montreal and Li Vaux Moise (al-Wueira) crowd the edge of their respective cliffs and are separated from the surrounding terrain by deep gorges.
Another method of keeping the trebuchet at arm's length was the use of natural or manmade ravines. At Saone a moat was dug that deserves a place on any list of historic engineering wonders. The castle sits on a long narrow triangle atop a rocky ridge. On the side where it touches the mountain the Crusaders hewed a deep channel, 60 feet wide and 450 ft long. The sheer, rock cut walls of this ditch rise 90 feet before touching the base of the castle. To create the moat, workers carved out an estimated 170,000 tons of solid rock then used it to make the walls and towers. No single span could cross the moat so a rock needle rises like an Egyptian obelisk from its center which once supported a narrow wooden bridge balanced atop it. Only via this bridge could anyone cross the hand hewn chasm. (A similar moat of equal depth was built at Kerak but has since been filled in). Eventually changes in attack methods led to the placement of cisterns on the inside of the outer wall where they became an integral part of the defense by discouraging sappers.
Other measures included the addition of a second set of outer walls. This gave rise to the so-called concentric castle. Strictly speaking few castles are concentric, i.e. possess a complete double set of curtain walls. The second wall was usually placed at the most vulnerable point of attack. The original designs placed a simple earth bank at the appropriate distance but this was prone to being captured by the enemy after which it became a platform for enemy artillery pieces. Later designs called for stone walls as large and as strong as the inner walls. The external walls increased in importance. By the 13th Century their fall or capture spelled the death knell for a garrison and its inevitable surrender since the loss of the outer wall meant that the enemy could place his artillery within range of the main castle.
Castle architects also responded by changing the shape of towers. The older square towers had two major disadvantages. First, their shape formed corners that were vulnerable to mining. Additionally, the flat faces of the tower offered inviting targets for siege artillery. A second major flaw was the considerable amount of "dead ground" in front of the outer face. This space could not be covered by archery from the curtain wall and provided the besieger with a comparatively safe place to begin mining operations or place siege engines.
Architects took different responses to this challenge. The Knights Templar favored oblong towers with the longer sides of the ellipse attached to the curtain wall. These towers projected sufficiently to allow archers and artillery to cover the ground in front of the wall, but not so far that they were vulnerable themselves to mining or artillery.
The second response was the round tower. The protective attributes of round towers were known by the Romans whose genius of fortification design, Vitruvius, had recommended their use in the 1st Century. Round towers have no vulnerable corners and stone missiles tended to hit a glancing rather than direct blow. But round towers were more costly to build and required skilled labor to erect them. The resulting internal round tower rooms were also less convenient to live in. Mounting siege engines on the roof was difficult since the available surface area is less. The earliest round towers were constructed at Saone.
Another change involved placing the castle's cisterns in front of or behind the curtain walls as a counter to mines by threatening to flood them. Siege engines and mining operations were also vulnerable to armed attack. As the principle of active defense took hold, castles were equipped with a number of sally ports -- small fortified gateways placed as posterns in the main gates and in the reentrant angles between the towers and the curtain walls. Small bands of the besieged could take advantage of lulls and gaps in the besieger's action, rush out, strike and withdraw. Several sieges were ended when the destruction of artillery or the loss of skilled miners disrupted attempts to reduce the castle.
Men Not Walls Make a Castle
These modifications, while not restoring the defense's superiority, were able for the most part to neutralize the advantages of the besieger that trebuchets and mining had brought. So how were castles taken? Why should garrisons protected by fortifications and provisioned like those like those at Kerak, Montreal or Saone ever have needed to surrender?
There is no single answer. It has always been possible to reduce strong points without ever putting their defenses to the test. A fortress can be blockaded and starved into submission. But as we have seen this was unlikely except in the case of the most ill-prepared defenses. It was also usually impractical since it required ample time and the absence of a relieving force bent on ending the siege -- conditions which were rarely satisfied.
But it was by these means that all a large number of cities and all the major castles of Oultre Jordain were captured by Saladin. What sealed their fate was his crushing victory of the Crusader field army at Hattin in 1177.
The collapse of these fortresses proved the old saying that a fortification is meant for the use of the garrison NOT vice versa, and that "men not walls make the city." Morale is a key factor to the defense of any fortification. If it can be destroyed then walls and towers cannot overcome its loss.
The defeat of the Crusader army at Hattin was complete. Castle defenders found themselves without hope of rescue by a friendly force able to challenge the besiegers and bring supplies. Their will to resist was broken. In the weeks that followed the garrisons of many strong places throughout the kingdom judged it useless to prolong resistance because there was no prospect that a relieving force would arrive in time. Saone, for all the strength of its natural and man-made defenses, surrendered to Saladin in three days. Tiberias, Jaffa, Caesarea and Ascalon gave up without a fight. The heavily fortified city of Acre yielded in a single day. Yet the following year it would hold against the Third Crusade for two years because Saladin had an army in the field that cooperated with the garrison. Jerusalem resisted long enough to get favorable terms (slavery rather than slaughter).
In the north the great castles held out and negotiated arrangements with Saladin who had overextended himself and was intelligent enough to recognize his limitations. But it was a stay of execution, not a reprieve. By the middle of the 13th Century, the increasing professionalism of the Muslim armies and the greater sophistication in siege artillery was joined by changes in mining procedures as surface sapping gave way to long distance tunneling.
The rise of Sultan Baybars coincided with a rekindling of the missionary zeal of Islam and the reorganization of Muslim energies for war. To the forces he commanded, the Crusader castles had become symbols of Frankish domination and infidel pollution. Each and every castle had to be taken and its defenders killed or expelled. Driving this end game was the belief by both sides that if they fell in battle they were assured of eternal life.
Baybars was a master of siege artillery and mining. Where Saladin had never used more than eight artillery pieces, the Mamluk Sultan routinely commanded up to 30. His effective use of both weapons meant that no Crusader castle was able to withstand a siege of more than six weeks.
In addition to military skill, Baybars also showed himself to be an expert of psychological warfare and deception. At Safad, he broke the unity of the defenders by promising quarter to anyone who was not a Frank. At Krac de Chevaliers he turned to other means.
When Saladin, fresh from his victory at Hattin, had reached Krac des Chevaliers he took one look at its defenses and decided that there were better uses for his army. Baybars conducted a relentless siege in 1265. For a month his men pushed in until they reached the fourth, inner line of defense which remained impregnable to artillery and mining. Under flag of truce a messenger arrived and met in the keep with the commander of the Hospitallers. The letter, signed by the Grand Master of the Order at Tripoli in Lebanon advised the defenders that there was no hope of raising a relief force and instructed them to negotiate a surrender. Baybars graciously allowed the Knight Hospitallers to withdraw under safe conduct in exchange for the fortress. Only after the warrior monks reached Tripoli did they discover the letter was a skillful forgery.
The Crusaders held on for a few more years, even briefly allying themselves with Genghis Khan in his rampage through the Levant. The last major castle, Margat, fell in 1285 to Sultan Qala'un in a manner that was both characteristic and fitting of the changes of the era.
Margat was the largest of the Crusader Castles and the last stronghold of the Knights Templar. Its concentric double line of black walls made it nothing less than impregnable. Its cellars were stocked with enough provisions to last a thousand men over five years. When Qala'un's siege weapons were destroyed by the defenders in a an artillery duel and sally that almost killed him, he brought up his miners. Cutting through the bedrock they dug a gallery that eventually undermined the foundations of the circular tower keep. When the work was finished the sultan, reluctant to destroy the structure, invited a delegation of the defenders to view the extent of the sapping work. Realizing that they were finished, the Templars surrendered on a promise of safe conduct to Acre and then to Europe.
The Crusader presence lasted a few more tense years until a handful of Christian adventurers robbed a caravan, hanged 19 Muslim merchants and then pillaged a nearby town. The Mamluk Sultan Khalil demanded satisfaction. Receiving none, he marched against Acre. After a siege of 43 days the city fell and Khalil allowed his men to massacre or enslave 60,000 prisoners. Tyre, Sidon, Haifa and Beirut fell in the weeks that followed. The last of the Franks were driven out.
But they had made their presence known. And everywhere they had once exerted power they left behind their fortresses and castles. Seven centuries later, the bravest of the stones still stand, square cut, fitted together with the fineness of a blade -- memorials to a long gone era.