The Accursed Tower

by Roger Crowley

  Khalil’s study of Acre’s defenses enabled him to site his largest catapults against the most promising objectives. These machines, which had been dragged from Cairo and across the Lebanese mountains, pulled out of the gates of Damascus on a tide of patriotic and religious fervor, dismantled again, hauled another eighty miles and reassembled outside the gates of Acre, represented not just a triumph of Mamluk military organization but a sizeable investment of resources and ergonomic capital. They were irreplaceable, and vulnerable to counter batteries and Greek fire. This was well understood, and they were positioned in line with al-Ansari’s advice: “He who has erected them for battling against a fortress should place them in a position which enemy [fire] cannot attain.”23 Each one took two days to erect, position, and prepare for action. They comprised the largest collection of stone-throwing machinery an Islamic army had yet assembled and the ultimate development of mechanical artillery in the crusader era.

  During this initial setting up, there was an eerie lull. From the Muslim camp, Baybars al-Mansuri noted an unnerving confidence in the defenders: “They showed great patience and a lack of concern about the siege and did not even close the gates of the city, nor even hang down a defensive screen [against bombardment] in front of the gates.”24 There were, evidently, forays and tentative probing by the defenders while the still-vulnerable camp was being established. The Templar of Tyre noted that the sultan’s army “sat before Acre for eight days, doing nothing beyond engaging in sporadic skirmishes between our men and theirs, in which a few died on each side.”25 The Islamic sources gave a more partisan account of something like a tournament between opposing horsemen: the Christians “began to go out to the Muslim military and call their knights to a duel. The troops from the free cavalry and the sultan’s Mamluks hastened toward them, and attack and retreat and mutual stabbing took place between them. They kept this up for some days, and the Muslims were victorious, and wounded a number of them and killed a number of them, and every day they returned the losers. They saw the Muslims had the heart they did not have. They ceased fighting and competing, and stood on the gates to protect them, and did not go out.”26 More likely, the Christians were simply outnumbered and common sense prevailed. The gates were sealed shut.

  9

  “BOLTS OF THUNDER, FLASHES OF LIGHTNING”

  April 10–13, 1291

  THE PROFICIENCY AT siege warfare that Khalil’s army brought to the walls of Acre, and the technologies that it could deploy, drew on traditions that went back to the very beginnings of Islam. Despite its origins in the deserts of Arabia, the taking of fortified places had been central to the spread of Muhammad’s message. The Prophet himself had besieged towns using trebuchets, a technology probably originating in China in the centuries BCE and transmitted to the Middle East via the Byzantines or the Persians. These machines utilized the energy generated by an off-balance beam rotating around a fulcrum to hurl heavy projectiles long distances. Trebuchets were the most powerful form of mechanical artillery ever devised. They had been an essential feature of Baybars’s siege trains; by the time of Qalawun, they were being deployed in mass batteries. “There were collected against Acre,” wrote Abu al-Fida, who had helped to transport them, “great and small mangonels such as never were collected against any other place.”1 Two years earlier, Tripoli had been taken with the aid of nineteen machines. The devices now being assembled under the gaze of those watching from the walls came to at least seventy-two and possibly as many as ninety-two. Their very presence was a triumph of Mamluk military planning and a testament to the depth of its resources.

  The power and variety of the machines that Khalil was preparing to unleash on the stone walls of Acre was impressive. They were of two types. In both, an off-balance beam would be unequally weighted around a pivot and supported on either a stout wooden trestle or a single pole. A sling would be attached to the longer portion of the beam containing the missile—a stone or an incendiary device—and then released. Initially, this was done by human energy. The original trebuchet was a traction machine, which depended on a team of men hauling smartly down on ropes in unison to flip the beam end upwards and release the missile from its sling. Baybars himself had participated in such launching work at the siege of Arsuf. By the end of the twelfth century, the ergonomic efficiency and power of trebuchets had undergone a step change with the creation of a second, larger device that employed the force of gravity to rotate the beam and hurl the projectile. Instead of using muscle power to fling the beam into the air, its short end could be loaded with a heavy counterweight—stones or lead weights in a sack or wooden box. The beam was longer, up to thirty feet, and the heavy wooden frame substantially braced with side supports to resist the greater forces at work. Further modifications included the lengthening of the sling containing the missile, which increased velocity and range, and in some machines a hinged counterweight that could swing from the beam, adding further impetus, in place of a fixed one. The loading and shooting rate of these machines was considerably slower than that of their traction-powered counterparts. A rope winch or human labor would crank the long end of the beam down to the ground, hoisting the counterweight at the other end into the air, and latching it into the firing position. A stone ball was lugged into the pouch of the sling placed within a wooden guide channel in the base of the machine and connected to the beam by a hook that would release the ball at the estimated optimum trajectory to hit a given target. With a release rope attached to the latch, the shooter would retreat a safe distance, give a firm tug to free the trigger, and the counterweight would crash downward with all the force of gravity, hurling the throwing arm skyward and releasing the ball toward the walls at frightening speeds, upwards of 130 feet per second.

  The operation of the counterweight trebuchet. This counterweight is a basket of stones hinged to the beam. In A, the counterweight has been cranked up into the air into the load position by a capstan. The sling with its projectile lies in the bed of the machine waiting the tug on the release rope. At B, the released weight hurtles to the ground with the force of gravity, and the beam flies up and fires the ball from the sling at the chosen trajectory. At C, the beam with its empty sling swings wildly to and fro in the air. (Sieges of the Middle Ages, London, 1964)

  The skill involved in setting up, range finding, and firing both types of machines was considerable, the result of a long evolution of practical experiments. There were many variables: the size of the stone ball, the length and flexibility of the beam, and the ratio of the relative length of the two parts around the pivot, as well as the length and adjustment of the sling. The faster-firing traction trebuchets called for different resources and skills. To create a continuous bombardment, large consignments of men worked in rotation in teams of up to ten at a time, hauling on ropes in perfect unison. The key figure was the shooter, who directed the team and prepared the ball for launching. This man held the ball in the sling prior to launch. He might be dangling off the ground, pulling the ball down, in order to flex the beam at the moment of launch, or standing on the ground and holding the sling and missile tightly to his chest. It was critical to position these at the correct angle if the ball were not to fly vertically upwards or not launch at all. The shooter required courage, knowledge, and fine judgment. He then gave the order to take up the slack and to haul. The beam whipped up, pulling the ball up out of his hands and releasing it from its sling at the chosen trajectory. The operation of trebuchets also called for considerable logistical and skilled manufacturing support: carpenters and fabricators; quarriers of stone; masons to carve projectiles; and the transport corps, which had hauled the machines to Acre.

  The siege of Acre a hundred years earlier had witnessed the power that these machines could unleash and the awe they inspired. Some of the largest had most likely been the relatively novel counterweight devices. According to the Christian chronicles, the Muslim defenders had

  plenty of trebuchets in the city, but one of them was unequalled for its mass
ive construction and its effectiveness and efficiency in hurling enormous stones. Nothing could stand against the power of this machine. It hurled huge stone-shot… If the stones met no obstruction when they fell, they sank a foot deep into the ground. This machine struck some of our trebuchets and smashed them to pieces or at least rendered them unusable. Its shots also destroyed many other siege machines, or broke off what it hit. It shot with such force, and its blows were so effective, that no material or substance could withstand the unbearable impact without damage, no matter how solid or well-built it was.2

  Fragment from a medieval manuscript showing the firing of a projectile from a traction trebuchet. The shooter suspended off the ground from the sling is using his weight to flex the beam and increase velocity and range. The men hidden behind the charging cavalry are awaiting the shooter’s order to haul downwards on the ropes and launch the ball out of his hands. (Sieges of the Middle Ages, London, 1964)

  The attackers in their turn deployed a machine christened “God’s Stone Thrower” that was positioned against the Accursed Tower and was said to have stripped thirty square feet from the top of its adjacent wall. Gravity had brought a considerable acceleration to the propulsion of projectiles even if these descriptions routinely exaggerated their destructive power. Over and above their actual capabilities, the spectacle of the loosed beams rearing wildly to the sky and swinging to and fro like a giant pendulum weighted by its stone box, the emptied sling whipping round, and the anticipated crash of the stone ball into walls, houses, and enemy camps had an immense psychological impact on defenders. Florid Arabic verses fancifully portrayed the trebuchets as instruments in the service of God, comparing the dip and rise of their beams to that of the faithful at prayer: “The prostrate catapults pray and we to God submit.”3

  Using the pulling power of a dozen men, a large traction trebuchet could probably throw an object of up to 110 pounds, though generally they were deployed to hurl smaller projectiles at a rapid rate of fire. The counterweight could hurl further and heavier. There was a trade-off between weight, trajectory, launch speed, and distance, and between all these and rate of fire. At Krak des Chevaliers, Baybars’s machines hurled stones of 220 pounds. At Montfort a few months later, stones of 155 pounds had been thrown one-eighth of a mile. Large machines would have required massive counterweights—the most powerful up to ten tons. Counterweight trebuchets also possessed a level of accuracy that made them extremely effective against static targets such as castle walls. They could pepper the same spot repeatedly. A medium-sized machine fired at a wall 200 yards away could reliably group its shots within a seven-yard square. It would be a long time before gunpowder artillery could overtake the effectiveness of these mechanical devices.

  Khalil came to Acre in 1291 with catapults of several types: lighter, quick-firing machines designed to rain missiles on defenders and heavier ones to batter walls. Arabic sources listed four different types. The ifrangi (the Frankish catapult), of which he had fifteen, was a heavy counterweight device for smashing battlements and the tops of curtain walls, mounted on a trestle frame, capable of hurling massive balls of up to 400 pounds, with counterweights of several tons. Then there were the somewhat lighter, man-operated traction devices—the qarabugha (the black bull) and the shaytani (the devilish)—probably also trestle mounted. One source lists fifty-two of the latter. There were also small antipersonnel devices on poles, which the Mamluks called the lu’bah (the plaything), capable of being swiveled in any direction to pepper defenders with small shot. In addition, some troops hurled stones with hand-held slingshots.

  The installation and management of the counterweight Frankish machines, with their sturdy frames and massive counterweights, was time consuming and laborious. The largest might take up to half an hour to crank up, load, and fire, and they could not be quickly re-sited. The judicious choice of objective at the outset was therefore important. The lighter, man-powered machines could be set up and repositioned more quickly. When a detachment of the Second Crusade besieged Muslim Lisbon in 1147, two traction trebuchets, each operated by squads of a hundred men working in rotation in ten-man teams, were reputed to have fired 5,000 stone shot in the space of ten hours—each machine hammering away at 250 shots an hour. At Acre, Khalil certainly had the human resources to unleash a heavy bombardment.

  The logistical demands of artillery warfare also required the supply of a vast number of suitable stone projectiles and the labor to quarry, transport, and shape them. From experience, the Mamluks understood the geological requirements of ballistic ammunition. The walls of Acre were constructed of the local kurkar, porous beach sandstone. This rock provided suitable ammunition for antipersonnel machines, but in order to degrade fortifications, it was far more effective to propel stone balls of a harder material than the walls at which they were aimed. Richard the Lionheart was said to have brought granite ammunition with him from Sicily for the first siege of Acre. To obtain rock of suitable density, Khalil sought limestone from other strata and transported it to the firing site. Harder beach rock was sourced from seams up to twelve miles from Acre. From this rock, skilled masons prepared well-formed spherical balls, probably turned on a lathe, in a range of sizes and weights to suit the requirements of various machines, up to giant stone balls of two feet in diameter, 360 pounds in weight. For the counterweight machines, supplies of equally sized spherical ammunition were essential to achieve consistent aerodynamic performance and accuracy.

  The Mamluks brought to the walls of Acre a wide range of other siege technologies and skills. These included a highly developed use of fire weapons by specially trained troops adept at the handling of Greek fire—a mixture of crude oil and powdered pine resin, which gave it adhesive properties—that could be projected in different ways. It could be hurled in clay pots from catapults to terrorize both troops and civilian populations and to destroy wooden siege machines and defenses or lobbed over walls by hand in small clay grenades. Specially adapted crossbows also propelled “eggs” of Greek fire and flaming arrows from a greater distance. It is possible that the Mamluks also brought genuine explosives into play at the walls of Acre; the author of a treatise describing the refining of potassium nitrate was living in Damascus at the time of the siege. To get up close to walls, the command of siege craft included the deployment of mantlets—moveable siege shelters—the watchful use of cavalry to deter sorties, and mining. The use of noise—chanting; martial music from drums, trumpets, and cymbals; and full-throated yells—was also a standard technique aimed at keeping defenders in a state of perpetual dread.

  Some Islamic sources recorded that it took two days to assemble the war machines out of reach of archers but within open view of the walls. This activity in itself was part of the Mamluks’ softening-up process; it was recommended that it “should not be concealed, because by doing it [openly], fright and terror and weakening of their hearts occur.”4 Just the sight of the giant trebuchets had the ability to drain the morale of a besieged stronghold: when King Edward I, erected an enormous machine christened the Warwolf before the gates of Stirling castle in 1304, the Scots tried to surrender before it had fired a single shot. (Having commissioned the giant machine, Edward was not going to let them off that lightly. He wanted to see it in action. He sent the delegation back inside so that he could witness for himself the destructive power of its stone missiles.)

  The Arabic-speaking Templar of Tyre watched the developments from the ramparts. He was evidently able to obtain some detailed information on the disposition of Khalil’s forces because he recorded not only the position of four particularly large counterweight trebuchets, some constructed from the tall trees of the Lebanon, but also their nicknames. The presence of each of these machines represented a vast amount of human labor, and in them were invested religious zeal and the expectation of victory. Near the northern end of the wall, in the section guarded by the Templars, was positioned an engine called “the Furious.” At the far southern end close to the sea, in the section g
uarded by the Pisans, was the similarly impressive “Victorious,” which Abu al-Fida had helped transport through the winter rains. Two more, which he did not name, were positioned against other strategic and vulnerable points—one against the Hospitallers’ section, where the wall took an inward turn, menaced the section close to the gate of St. Anthony, the main thoroughfare into the city, another against the exposed protruding right-angled turn, protected by an external barbican, which Khalil had identified as the most promising point of attack, and whose inner defense, the Accursed Tower, gave direct access to the heart of the city.

  The spread of these machines from shore to shore suggests Khalil’s desire to keep the defenders stretched along the whole perimeter, even if it was in the central sections that his deeper plans lay. The defenders had an unknown but smaller number of trebuchets of their own, the construction and operation of which was largely in the hands of the Pisans. They evidently had some “great machines,” counterweight trebuchets that were probably deployed throughout the city within the inner wall to target Khalil’s own counterweight monsters, but as the siege progressed, they would run short of suitable ammunition.

  By sometime around April 11, the Mamluks’ trebuchets were assembled and in position: the stone balls collected in piles, the pulling teams ready, the counterweight machines sited and cranked to fire. Khalil’s strategy was to move fast, not to give the defenders a moment to breathe. “When the investment is under way,” it was advised in the manual of al-Ansari, “there should be no pause in the discharging of the mangonels against them, and there should be no abating [of fire] in any hour of the day or night. To desist in attack against them is among that which cools fright and strengthens their hearts.”5 While the catapults unleashed a withering torrent of fire, his troops inched forward toward the outer ditch, by night and day, demolishing any outer works in the process. With growing apprehension, the Templar of Tyre watched their method, conducted with extraordinary discipline. Their progress seemed unstoppable: