"Secrets of Lost Empires: Medieval Siege"

PBS Airdate: February 1, 2000
Go to the companion Web site

SPONSOR: During the following program, look for NOVA's Web markers which lead you to more information at our Website.

NARRATOR: Knights in shining armor charging each other in great pitched battles - that is the popular view of warfare in the Middle Ages. In reality, most medieval conflict involved the attacking of castles, and mounted knights were not much use against stone walls. A siege against a castle could take weeks, even months - the attacking army pitted against a well-defended garrison within. By the end of the 13th century, the science of defensive architecture had reached a peak. Stone walls were built thicker and taller than ever before. And archers could easily pick off advancing attackers. To further frustrate attempts at breaching the walls, castles were situated on rocky crags, or surrounded by water. But every advance in castle defense drove attackers to devise better siege weapons.

RICHARD HOLMES: During the Middle Ages castles kept improving. They kept improving as weapons of attack got better. And tactics was this eternal balance between attack and defense.

NARRATOR: 200 years before cannon appeared in Europe, chroniclers make reference to what appears to be the ultimate 13th century siege weapon - an ingenious new form of heavy artillery that flung huge stone balls with such destructive power that castle walls were reduced to rubble. But no ancient weapon of this type has survived. Were such claims gross exaggerations, or did such a weapon really exist? To answer these questions, NOVA brings together a team of experts in medieval warfare who believe they know the secret.

JOEL MCCARTY: It's chaos. It's Wednesday, I think. I don't have a clue whether or not we'll finish.

NARRATOR: Their task - to build siege machines capable of destroying a castle wall at a range of about 200 yards.

HEW KENNEDY: I think that the thing'll smash it up nicely, yes.

NARRATOR: With just two weeks to meet the challenge, a siege mentality quickly sets in. No modern builders have ever managed to do this before.

TIMBERFRAMER: And the whole thing could twist and kick quite alarmingly.

MARCUS BRANDT: This interplay between defenders and siegers, it's still up in the air. We could still take it, then again, we could fail. It's sort of in the lap of the gods.

SPONSOR: Major funding for NOVA is provided by the Park Foundation, dedicated to education and quality television.

SPONSOR: CNET.com, helping you choose the right technology product.

SPONSOR: This program is funded in part by Northwestern Mutual Life, which has been protecting families and businesses for generations. Have you heard from the quiet company? Northwestern Mutual Life.

SPONSOR: Additional funding for this program is also provided by the David H. Koch Charitable Foundation. And by the Corporation for Public Broadcasting and by contributions to your PBS station from viewers like you. Thank you.

NARRATOR: In the year 1304, Edward Longshanks, more formally known as King Edward I of England, mounted the greatest siege of his reign against the Scots and their castle at Stirling. The attack dragged on. Impatient for victory, Edward ordered 50 carpenters to immediately begin building a monstrous new weapon - so powerful it would breach the strong walls of Stirling Castle. Details about the weapon design are tantalizingly vague, except that it was nicknamed Warwolf, and its appearance outside the walls terrified the garrison. Was it the atomic bomb of the Middle Ages? With one blow, Warwolf leveled a section of wall, successfully concluding the siege of Stirling Castle. What kind of a weapon was Warwolf? Hew Kennedy is a Shropshire landowner and medieval armor expert. 10 years ago, he became intrigued by a picture of a machine drawn by Leonardo da Vinci. It appeared to be a device for throwing dead horses, called a trebuchet. Inspired by the power of a machine that could hurl such heavy missiles, Hew decided to try building one himself—a quest which eventually led to this piano-flinging contraption, a mechanized catapult made from a laminated beam, scrap metal, telephone poles and steel cable. In essence, a trebuchet is a giant seesaw with a very heavy weight at one end and a much lighter missile attached to the other. As the heavier weight drops, the lighter projectile is whipped by its sling towards the enemy. Hew is convinced that Warwolf, Edward's great wall busting siege engine, must have been a trebuchet.

HEW KENNEDY: If you chuck a thing that heavy at a stone wall it'll shatter it. Stone missiles are a lot more effective than grand pianos.

NARRATOR: To test out Hew's confidence in the destructive capability of a medieval trebuchet, NOVA is preparing some hard sandstone balls weighing 250 pounds - and a wall. It's made of sandstone and lime mortar. In construction and design, it is based on the upper section of a typical castle wall of the 13th century. Hew wants to build a trebuchet capable of knocking it down, but at a range of 200 yards, it will require precision as well as brute force. Michael Prestwich, a medieval historian, will ensure that Hew's next trebuchet will be based on an authentic 13th century design.

MICHAEL PRESTWICH: And I suppose when its got to the top of its trajectory, it starts coming down again. It really looks quite frightening. And I'm glad I wasn't standing underneath it.

HEW KENNEDY: It would bust up a building all right, wouldn't it?

MICHAEL PRESTWICH: It's the first time I've seen a full-scale trebuchet in operation. To see the high trajectory of it and the way the missile and the sheer speed with which it falls, it is a fantastic sight. Trebuchets began in the Far East, in China. But what they were there were hand-pulled machines worked by quite large teams of men.

PAUL DENNY: Prepare to loose, loose.

MICHAEL PRESTWICH: In many ways, quite they are limited in what they could do. The big advance came when Arab engineers got hold of these devices and put a big counterweight on so that instead of teams of men pulling it, the beam was pulled down by a great counterweight. They were far more potent and far more effective. These machines were picked up by western engineers, and by the middle of the 13th century, it's very clear that French, English engineers were capable of building really quite large machines.

NARRATOR: Some of the best military engineers were employed by Edward I - a master of military tactics. He was one of the most vicious and single-minded rulers of his time. Soon after ascending the throne in 1274, Edward decided to squash Welsh independence and bring Wales under his personal rule.

MICHAEL PRESTWICH: He was a bully, frankly, and I think many people would think of him as a really nasty piece of work. He was utterly determined. Nothing was going to get in his way.

NARRATOR: Edward's strategy was to ring the mountain stronghold of the Welsh prince with a chain of powerful castles. Richard Holmes is an historian of military tactics.

RICHARD HOLMES: He built eight new big castles, which were really state of the art. They were immensely strong, well thought out. And most of them could be supplied by water so they were very difficult for the Welsh to besiege. And Edward believed that you controlled the countryside by castles like this. They're like nails holding the landscape down. And their garrisons could issue out, attack enemies in the area. And until the castle was taken, nobody could really dominate that landscape. They were extraordinarily expensive to build, and were a very severe drain on the royal exchequer. In the short term, though, they worked.

NARRATOR: Edward and other English lords designed their Welsh strongholds with the trebuchet in mind. For example, Caerphilly Castle was surrounded by manmade lakes which kept a besieging army and their siege weapons at a distance.

RICHARD HOLMES: Castles were what modern tactitioners would call force multipliers. They enabled a relatively small garrison to operate at the absolute maximum of effectiveness. And a castle like this is carefully organized to maximize defensive firepower. There are loopholes in the walls and the towers for archers to shoot through. And here the walls are cunningly organized so that the second set of walls is higher than the first. And therefore an attacker facing this face of the castle not only gets the defensive fire of the first wall, but he's got archers shooting at him from the higher walls behind it. It's a real nightmare.

NARRATOR: At the end of the 13th century, what was the effective range of an archer? And what was the effective range of a trebuchet? The historical reports differ.

RICHARD HOLMES: Hew, how close are you going to have to bring your trebuchet to the walls to do serious damage, do you think?

HEW KENNEDY: Probably 200 yards, we will need to be within that to smash it up.

NARRATOR: At 200 yards, is Hew's trebuchet out of range of archers defending the castle? To find out, a dummy representing the trebuchet's chief operator is placed at that distance.

HEW KENNEDY: I'm sure an arrow would land amongst us if we're at that range. You could easily shoot 200 yards with that massive bow of yours, couldn't you?

SIMON: Yeah, 300 yards.

HEW KENNEDY: Yes. Well at 200 yards, I think it would be putting you a bit worried, wouldn't it?

RICHARD HOLMES: Yes, it would. I'm the first to accept that from this sort of range the trebuchet would be doing serious damage to the castle walls. But I think this does suggest that it's no easy business. And the garrison that knows its business can probably keep a trebuchet at the very limit of its range. And the fact that we had some going over the top I think is mighty hopeful from the archer's point of view.

HEW KENNEDY: I wonder what happens if you slap one into him from here? Come on then. It's all right, he's swallowed it, hasn't he?

RICHARD HOLMES: Gone right through the dummy, kept only in by the fletchings.

HEW KENNEDY: Bit of a bellyacher I reckon, yeah.

NARRATOR: Edward's castle-building campaign in Wales had taught him how to design well defended fortresses. Turning his attention to conquering Scotland, did Edward also have the ability to successfully attack them? As the king marched northwards to take the castles that guarded Scotland, he brought with him some of the biggest siege engines, or trebuchets, ever built.

MICHAEL PRESTWICH: The siege of Caerlaverock, conducted by Edward I in 1300, we've got remarkably a really good account of this in a contemporary poem. It describes the way in which the knights rode up to the castle all in their great armor, trying to perform great deeds of valor. In fact, they were driven back by the garrison, hurling stones and such at them. And it wasn't the knights, it wasn't these people with the great acts of bravery. It was the engineers, men of really quite low social status in comparison, with the great siege engines. It was they who compelled the garrisons to surrender. And the poem describes the way in which the great boulders came down from the sky, into the courtyard, crashing down, causing all sorts of damage and mayhem inside. The minute the casualties started, the garrison simply surrendered. So it wasn't the knights, it wasn't a great act of chivalry to capture this castle. It was the work of the experts, the engineers.

HEW KENNEDY: It's difficult to tell, I mean that one's obviously got a . . .

NARRATOR: Joining Hew Kennedy in his quest to build a trebuchet is mechanical engineer Wayne Neel, a professor from Virginia Military Institute.

WAYNE NEEL: This one actually is one to two, and this is one to three.

NARRATOR: Wayne will design the trebuchet. He is basing it on a picture he found in a 13th century Spanish manuscript. The illustration gives no idea of the true scale of the trebuchet because the artist has made the machine smaller than the soldiers standing next to it.

HEW KENNEDY: If the drawing isn't practical from all points of view, you begin to wonder about all the other points of view. It may be that the things they have got in the right proportion they did by accident. Even an artist like Leonardo will draw a plan for something that is totally impractical, but it indicates how it could be made. That's all. It indicates how it could be made.

NARRATOR: With the manuscript as a starting point, Wayne uses a combination of engineering theory and trial and error to come up with a working model. The prototype looks promising to Marcus Brandt, a carpenter who will help Wayne build the trebuchet.

But there's a problem. The model rocks dangerously. This would be a serious flaw in a full-size trebuchet with several tons in motion.

WAYNE NEEL: The machine basically tends to want to tip forward as that weight comes down. It wants to drop straight down, so it tends to pull the machine forward making it in this case tip.

NARRATOR: To resolve the problem, Wayne re-examines the medieval illustrations of this type of trebuchet. He's struck by the fact that many have wheels, so he decides to build a new model. Not only do wheels eliminate the tendency for it to tip over, the movement also boosts the trebuchet's performance.

HEW KENNEDY: Shoot away, come on.

NARRATOR: Hew doubts that wheels will make a trebuchet throw farther, so he insists on a demonstration. For the first throw, Hew holds down the model. The projectile travels 20 feet.

HEW KENNEDY: Now let's try it with - letting it loose. Well, it's certainly better, isn't it?

NARRATOR: When the trebuchet is allowed to roll, the missile goes an extra 10 feet. Why? The falling counterweight drives the trebuchet forward. Like a pitcher stepping forward, this adds momentum to the throw. The forward motion also permits the counterweight to drop further in a straighter line. The closer the counterweight follows this optimal path, the more energy it captures for throwing.

HEW KENNEDY: What I find odd is the idea about it moving and having to move in order to give more energy to the missile. I would have thought this was - on the face of it, nonsense. But obviously if Wayne's tried it, and it goes further, then it can't be argued with.

NARRATOR: Now that Wayne has confidence in his model, the project moves to Loch Ness in the Scottish Highlands for a full-scale test. The trials will take place in the shadow of Castle Urquhart, which may have been besieged with similar weapons during Edward I's Scottish campaigns 700 years ago. Wayne and Hew arrive at Castle Urquhart to inspect the timber that is delivered by barge. As in the Middle Ages, heavy duty English oak is the choice for building the carriage and trestles. But the throwing arm will be made from a more lightweight wood, Douglas fir. Wayne calculates that a tree trunk of at least two feet in diameter is needed to withstand the stresses of hurling 250 pound balls.

JOEL MCCARTY: How about this one?

WAYNE NEEL: It's the right size. That would be the right diameter.

TIMBERFRAMER: Be a shame to cut that down, wouldn't it?

WAYNE NEEL: Yes.

HEW KENNEDY: Let's get the chain saw and whack it down.

MARCUS BRANDT: Haul away.

NARRATOR: After it is floated across Loch Ness, the log is quickly hauled ashore. Work immediately begins on hewing it into an eight-sided throwing arm. The next job is to assemble the base. 40 carpenters, mainly from the United States, but also from Britain and Germany, have volunteered to spend their vacation here. They're all timber framers who specialize in traditional construction techniques. Without nails, they connect large pieces of wood using mortise and tenon joints. While the timber framers immerse themselves in medieval methods, Wayne employs the tools of a modern engineer to check and recheck his design. In particular, he's hoping that the wheels will give him the same boost in performance as they did on the model. Up at the target wall, work is almost complete. The wooden structures on top are called hoardings. They provided additional protection for defenders during the siege. Attackers would build similar barriers to protect the trebuchet team from arrow fire.

MICHAEL PRESTWICH: Well, this is the target wall that we've got for the trebuchets - a modern reproduction of a castle wall. It's a pretty good reproduction I think, it's got all the details of the crenelations. And what's important about it is the width. It's a good five foot thick. The way that it's formed is that there's an outer skin on either side, and within that a fill of rubble and mortar. And that provides a very solid core. This is exactly how they did it in the Middle Ages. And I'm very glad that I'm not standing here while the trebuchets are actually shooting.

NARRATOR: The target wall is modeled on the outer walls of Harlech and Caerphilly, two of the best defended castles in Wales. Faced with the daunting task of taking a castle like Caerphilly, what siege tactics would an attacker adopt?

RICHARD HOLMES: A castle like Caerphilly, like this one, presented an attacker with a knotty problem because it's got layers of defenses, with sandwiches of water between them. A bit like the layers of an onion. So an attacker arriving here would encounter the moat, a water defense, so that he couldn't mine underneath it without his trenches being flooded. His best bet if he possibly can is to take the place by surprise or by guile. To trick his way in - to get in before the garrison's ready for him. If he can't do that, he's then got to mount a formal siege. And ultimately he'll starve it out. Gates had once been the castle's weakest point. But by this stage in history, they'd become its strongest. They're defended by gatehouses, these are very powerful towers with arrow slits in them. So anyone attacking the gate is subjected to close-range fire. And there's a portcullis which drops down just behind the attackers, and even if an attacker does manage to take the gatehouse, then he's got another moat, two more gatehouses, and he's got to do the whole rotten business all over again.

NARRATOR: Meanwhile, rotten weather has besieged Castle Urquhart. And after a week of daily downpours, the trestles that support the throwing arm are ready to be raised. At the top of the trestle, the throwing arm rotates on an axle that has to be both strong and exact. Marcus Brandt has taken on the job.

MARCUS BRANDT: Well, we've got to address the issues of axles. We've got two sets of axles with this fixed arm, fixed counterweight machine. We've got the axles down here to carry the wheels, and we've got this more precise axle up here which carries this eight tons of weight and arm. And it has to be fairly precise. We've set up a great wheel lathe which is a precursor to modern lathes. It's just a great big flywheel. It's powered by muscle and it turns this 10" x 10" down to an 8" x 8".

NARRATOR: Marcus restores old buildings in Pennsylvania, and has enthusiastically embraced the idea of using medieval technology.

MARCUS BRANDT: It's a very medieval feel to the whole project. We've got all the trades going at once. We have the stone masons busy, we've got all the carpenters and the axes flying and the chips and the smoke and the mud. Boy, if it weren't for the jets flying overhead occasionally, you'd think you're in the 12th century.

NARRATOR: Villages were often established next to the castle of a nobleman or the king.

RICHARD HOLMES: English society was very hierarchical, with the king and his nobles at the top, and then the rest of society cascading down from that. Now the castle wasn't just important militarily, they were an economic power as well. There were communities built up round the castle to supply it. And the castle provided protection as well. At their best these noblemen were protectors, at their worst they were often something rather like mafiosi, leeching off the local countryside. So living under the shadow of a place like this had advantages, but disadvantages too.

NARRATOR: In order to position the heavy throwing arm, the timberframers place an A-frame above the trestles to support the pulley system - a standard medieval device. A block and tackle dramatically reduces the number of people required to pull on the ropes.

MARCUS BRANDT: Now block and tackle is a fairly simple device. It magnifies your pull. If I were to lift my own body weight with a single rope, I would have to pull down 200 pounds to lift my 200 pounds of body weight up. With this, I'm hooked up to this double sheath pulley, and I've got four lines which share the load equally, so 200 pounds here is only 50 pounds on each one of these ropes. So to lift my 200 pounds up, I only have to pull 50 pounds here. Now mind you, I have to pull four times as much rope, but it really works.

NARRATOR: The next phase of construction is the most dangerous. Without the help of a modern crane, the timber framers must trust the strength of two slender poles and the rigging of their ropes in order to raise the one ton throwing arm.

JOEL MCCARTY: One accident, one slip, one failure hidden in the heart of this timber and we're out of business - simple as that.

WAYNE NEEL: It's looking real good right now, going right on in there like we planned.

TIMBERFRAMER: We are in place.

NARRATOR: With the throwing arm in position, Wayne needs to attach weights that will power the machine. He chooses lead because he believes lead was used for the counterweight of Warwolf, the mysterious and terrifying siege engine that clinched the assault on Stirling Castle. There is good historical evidence for this. In a letter Edward wrote just before the siege, he demanded that lead be removed from all the churches.

MICHAEL PRESTWICH: One of the things that they needed for the siege of Stirling was heavy weights for the counterweights on the engines. And they sent orders out to strip all of the church roofs in the entire surrounding area. So all of this lead, lead sheets, would have been brought to the siege and then melted down in order to form the counterweights.

NARRATOR: Wayne's calculations tell him that in order for his machine to throw a 250 pound ball, he'll need about six and a half tons of lead counterweight. It takes a week to melt down the scrap and form it into a collar that can be bolted onto the arm.

MICHAEL PRESTWICH: Wayne's come up with a really neat solution for a lead counterweight made from various pieces bolted together. But it is expensive. Lead is difficult to get hold of. It's complex to make this. There is a much easier solution, which is to use a large box as a counterweight. Something like this, which you could fill with earth, with stones, with anything you'd got. Fill it up. You could put as much or as little in as you wanted, and that would affect the range of the machine. And you'd simply then have to put this box on the machine, swiveling roughly like that.

NARRATOR: A survey of medieval illustrations suggests that swinging counterweights were the more popular design. A man who has been working with them longer than anyone else is traditional French carpenter Renaud Beffeyte. Renaud is convinced that Edward's great siege engine, Warwolf, had a swinging counterweight box. The simple design of the box makes it not only cheaper, but easier to build than one with a lead counterweight. And Renaud has found graphic accounts of the destructive power of some really big siege machines.

RENAUD BEFFEYTE: I found account medieval -

MICHAEL PRESTWICH: An account, a medieval account.

RENAUD BEFFEYTE: Yeah, an account. And they pay 300 bullets in three days, then they shoot 300 bullets on the same place on the same wall. After three days, the wall was destroyed.

MICHAEL PRESTWICH: So you think the real purpose of something like this is simply to attack a castle from a distance and bring the wall down?

RENAUD BEFFEYTE: Yes.

NARRATOR: 500 years before Newton's apple, medieval engineers had figured that trebuchets with a swinging counterweight are the most efficient at using the force of gravity. Like adding wheels to a fixed-counterweight trebuchet, the hinge allows the swinging counterweight to descend further in a straighter line, capturing more energy for the throw. 15 years ago, Renaud came across the notebook of the 13th century French architect, Villard de Honnecourt. In it he found the plan for the base of a trebuchet and a description of the swinging counterweight box. Unfortunately, the page showing the rest of the trebuchet was missing. But with his knowledge of medieval carpentry and his experience of building over 30 small trebuchets, Renaud realized that if the design were ever built, it would be monstrous - the height of a five-story building. Ever since this discovery, Renaud has wanted to construct the full-size trebuchet. Finally, he will get the chance at Castle Urquhart in Scotland.

At the start of a medieval siege, a trebuchet engineer would typically seek out as big a throwing arm as possible. Renaud inspects a good-sized oak log that he hopes will be suitable. Using the same geometric principles employed by the medieval engineers who built castles, cathedrals and siege machines, Renaud starts designing his trebuchet. The key decision is where on the throwing arm to position the main pivot point, or fulcrum. Medieval engineers worked this out by observation. In this largely illiterate society, carpenters used animal figures like these as an aid to remember geometric formulas. Now that Renaud knows where to put the main axle, he can design the trestle, base and capstans to raise the counterweight box - because they are all sized in proportion to the throwing arm. But just as he begins building the trebuchet, Renaud is called back to France. In his absence, work progresses on the base. Work also begins on drilling out the main axle hole in the throwing arm. Unfortunately, there is some confusion about the size of the opening, and a nasty surprise awaits Renaud's return.

HENRY RUSSELL: Renaud's been away for a couple of days, and while he was away, we cut the axle hole. It's quite apparent that a lot of meat has gone from the timber and we're left with quite thin sections on either side.

RENAUD BEFFEYTE: I think the beam, the throwing arm, can break here, in this part.

MARCUS BRANDT: Renaud's concern is that we've gone too much to this extreme where we've taken out too much wood of the throwing arm and made the axle too strong, and his concern of course is that on our first throw, that it's going to do something like that.

NARRATOR: Renaud is left with no alternative but to hope that a couple of planks will be strong enough to splint the weakened throwing arm. Finally, work starts on raising the great trestles. With each passing day, the situation at Castle Urquhart more and more resembles a siege of old. With two trebuchets to finish, there is a nagging feeling that Renaud's machine may not get done before the timberframers must return home. Over at Wayne's trebuchet, work on attaching the lead weights is finally finished. All that remains is to cock the arm. Even with the help of pulleys, this requires 40 people. With a 13,000 pound counterweight, it's a much bigger job than any one imagined. And it will have to be repeated for each fling. The trigger mechanism is not strong enough, and buckles alarmingly as it takes the full weight of the throwing arm.

TIMBER FRAMER: It's bending as we're releasing the arm, and we're wondering whether it can hold the weight right now.

NARRATOR: To avoid an accidental firing, they reinforce it with a length of chain.

HEW KENNEDY: They've got lots of engineers on this and maybe they've got it right, but I think it's quite difficult to get it right.

JOEL MCCARTY: It's chaos, the sun is going down, I don't have a clue whether or not we'll finish even one of these machines on time.

MARCUS BRANDT: Well, we're under the gun now. It's kind of touch and go whether we're gonna get this thing put together and actually fling both machines, but if I know this crew, come hell or high water, we're gonna make this thing fling. There goes your light (laughter).

NARRATOR: The next morning a 250 pound sandstone ball is quickly positioned in the sling for the first throw.

HEW KENNEDY: It's got about a fifty percent chance of going in the right direction. It could go in the lake there by mistake. Because these adjustments are not well known to us.

TIMBER FRAMER: You ready? This is it. One, two, three, fire in the hole!

NARRATOR: Even with 12 timber framers pulling, the firing pin won't budge.

HEW KENNEDY: It's quite normal that, I mean, triggers are very stiff. We found that at home. It takes a lot of effort, which is as well, in some ways.

NARRATOR: Despite Hew's predictions that a trebuchet on wheels would shake itself to pieces, the reverse is true. Wheels dampen the recoil. Everything appears to be working as Wayne predicted, except the range. The ball only traveled about 170 yards, falling short of the wall.

GRIGG MULLEN: What are we . . .30 yards short, and about 5 feet low. I thought the throw was flat, which means we've got to - I think shorten the sling so it releases sooner and goes higher, which should get us to the wall.

NARRATOR: Just like his medieval counterpart, Wayne uses a process of trial and error to alter the trebuchet's range by adjusting the length of the sling. With the first throw, a long sling resulted in a late release and a low trajectory. By shortening the sling, Wayne believes the ball will be released earlier, resulting in a higher path. Wayne's adjustments have the desired effect. The second throw has perfect range, just missing the target to the right by two feet.

____: I think we took the King's ear off.

____: Good job, Wayne!

____: So it needs a slight adjustment, and then it would be a direct hit.

____: Tomorrow is another day.

NARRATOR: What will be the last day for Wayne and his team dawns, with the wall still intact, but clearly under threat. Meanwhile Renaud's trebuchet is still not finished.

This morning they rush to complete the counterweight box that is hinged to the throwing arm.

MARCUS BRANDT: We've got the one treb built; the other one is about 99% of the way there. We flung two stones last night with Wayne's treb, but we really don't want to let Renaud down, and we don't want to let ourselves down. We really want to see this thing fly before we go.

NARRATOR: Still, with time so short, it's unclear that Renaud will get a chance to fire his trebuchet at all. After last night's narrow miss, Wayne's trebuchet is repositioned to be more in line with the target.

WAYNE NEEL: All we're doing is shifting it slightly to the left. We were throwing to the right a little far. So we've shifted it about one inch, so that hopefully we'll be dead on center.

NARRATOR: With the same 250 pound ball as yesterday, and the sling at the same length, Wayne believes he is now dead on to hit the wall with his third attempt.

TIMBERFRAMERS: Five, four, three, two, one. Fire in the hole!

NARRATOR: The third shot is identical to the second in distance. At a range of 200 yards, adjusting the wheels one inch to the left placed the missile bang on top of the hoarding.

WAYNE NEEL: We've gotten wood, I don't know if we've contacted any stone yet. But we've knocked the hoarding pretty well. It was maybe a little high, and we've come down a little bit.

HEW KENNEDY: That must be the old hole there, mustn't it?

____: That's the hole.

HEW KENNEDY: It's still a bit on this side, isn't it? If you'd been standing under that hoarding, you'd have had a jolt.

NARRATOR: With the trebuchet lined up on the target, Wayne only has to shorten the range by a hair to hit the stone battlements below.

WAYNE NEEL: So this time what we've done is lengthen the sling about six inches, so we're hoping to fire a little bit flatter and get to the top of the hole - the top of the wall.

HEW KENNEDY: It's quite difficult, I think probably, isn't it, on those small adjustments? I mean, do you think -

WAYNE NEEL: Yes, small adjustments are difficult to do.

HEW KENNEDY: We could perhaps not get so lucky this time, but it's been very good so far, hasn't it?

WAYNE NEEL: Well, we'll see.

NARRATOR: Worried that this may be the last attempt, Wayne makes a sudden change of plan. He replaces the 250 pound ball that he's been using with a jumbo 300 pounder. Wayne figures that the heavier ball, clocked at a speed of 127 miles an hour, should breach the wall. But he's wrong.

WAYNE NEEL: We're going back to the 250 pound ball, instead of the 300 to see if we can get a little bit more height.

HEW KENNEDY: They've only got time for one more shot, the American team, and they've been very near, but they might miss it again. And if they don't get it, do you think you can get it with yours?

RENAUD BEFFEYTE: If we are more lucky, then we can destroy this wall.

HEW KENNEDY: Yeah. Well, we'll see. It's French against Americans.

RENAUD BEFFEYTE: No, it's - no, no, it's not against. That's your way.

NARRATOR: Putting on their kilts for good luck, Wayne's team rushes to get in one final shot.

TIMBERFRAMER: We did it! Come and take a look at the rocks over here. It just pulverized the stone on the inside.

HEW KENNEDY: It confirms what we came here to prove, didn't it? We - they've had a lovely hit, smack in the middle, and it smashed it. And it's busted it right through to the back. So it's quite obvious that if you've got one of these trebuchets and you've got a castle like this, and you've got plenty of time to shoot it, you're going to knock it into a powder. We can reduce this to rubble.

NARRATOR: Intoxicated with success, the timberframers bid adieu to the Highlands. But the next morning, Renaud is heartened to find that he's not been completely abandoned. Ed Levin and a handful of the Americans have decided to stay on in Scotland to help finish the job. The biggest concern is whether the throwing arm has been fatally weakened at the point where the main axle passes through it. To avoid stressing the arm, Renaud decides to only partially load his counterweight, using four tons of sand in the 12 ton capacity box. But there are risks to this approach.

ED LEVIN: We're all suitably cautious in having the 250 pound sandstone ball end up in the castle wall rather than in the loch, or drop down on the machine, or any of the other places it has historically been known to go.

HEW KENNEDY: Nobody knows quite what they're doing, so that's what makes it fun.

ED LEVIN: A moment of birth.

TIMBERFRAMER: And terror.

ED LEVIN: Yeah. Well, birth is usually accompanied by terror.

RENAUD BEFFEYTE: Now we are ready for shoot. It's getting me nervous.

TIMBERFRAMERS: Four, three, two, one, fire in the hole!

NARRATOR: The heavy ball and relatively light counterweight result in the missile landing dangerously close to the trebuchet.

HEW KENNEDY: Well yeah, I mean, we knew that there wasn't enough weight in really, didn't we? It was just an experiment.

RENAUD BEFFEYTE: The counterweight is not so heavy. We must put two bag more. Two tons.

HEW KENNEDY: Of sand.

RENAUD BEFFEYTE: Yes, of sand, yes.

NARRATOR: Two more tons of sand are added.

HEW KENNEDY: I don't think there's enough weight for it to go really well yet. This machine wants a lot of weight. 10, 12 tons probably, to make it go properly. Renaud thinks if we keep putting little bits in, he might just get there without busting the axle, which is natural, of course, because it's his machine. Fair enough.

RENAUD BEFFEYTE: We are going to get a good shot. I'm sure, sure, sure.

TIMBERFRAMERS: Five, four, three, two, one, fire in the hole!

NARRATOR: Renaud's optimism is justified. The missile falls just a few yards short of the wall and a bit to the right. The team decides to give it one more day. But the next morning starts with snow, followed by a heavy downpour.

MARCUS BRANDT: Well, basically we're at the last day, we've got - between the rain and the mud, we've got a rigger's nightmare. It's really taken its toll on the ropes, the mud grinds in and it starts tearing up the fibers. The water helps make the rope stretch, and if you look around the place, there's ropes in the mud and no rigger likes seeing that. So we're doing the best we can to keep our ropes clean, but it's an uphill battle.

NARRATOR: Last night's final shot was short of the wall because it was thrown too high. Renaud believes the sling is slipping off its prong too soon. So to delay release and lower the trajectory, the prong is bent forward.

ED LEVIN: We've got the right amount of loft, we've got the right amount of range, we're just missing the target off to the side.

NARRATOR: For days, Renaud has suspected that his trebuchet is pointing just to the right of the wall. But the loaded machine is too heavy to shift, and he faces thepossibility that he may have to go home having achieved only a near miss. At the last minute, Marcus offers a solution.

MARCUS BRANDT: If we had our preferences, we'd be able to move the machine over a little bit, but we're afraid of shattering the machine, particularly with all the weight in the basket. So we're gonna move the channel of the ball a little bit to the side so we can change our angle of attack. The range is good, but we just want shift it over to the left a bit.

NARRATOR: By shifting the channel that holds the ball slightly to the left, they hope to redirect the missile. It works - almost. Another three feet and Renaud would've had a direct hit. Unless shifting the channel was just a fluke, one more nudge to the left should bring the trebuchet right on target.

____: Go, baby!

____: Oh, that looks good!

____: Yeah!

NARRATOR: After two throws which are slightly high of the wall itself, Renaud orders a minute adjustment of the prong in order to lower the trajectory. With frayed ropes and a storm threatening to close down the siege, everything now hangs on Renaud's ability to quickly get on target - and he does. A bull's eye on the battlements.

____: Well done!

MARCUS BRANDT: This whole wall, if you run your eye down here, it's bellied out, there's cracks all through it. Anybody standing back here would've been mincemeat.

NARRATOR: In a real siege it would only be a matter of time before the wall is reduced to smithereens.

MICHAEL PRESTWICH: In terms of the kind of dialogue that existed between attack and defense, it is very clear now to me that the appearance of the trebuchet on the scene shifted that balance radically in favor of attack.

MARCUS BRANDT: I've gained tremendous respect for the medieval engineers. They were able to build a frightfully powerful and highly accurate and easily adjustable machine. If you're under siege, you've got to try to knock these things out before they're actually built, cause once they're built, you're sunk.

RENAUD BEFFEYTE: The trebuchet is this big machine who can broke the wall and also a trebuchet must be the Wolfwar.

____: Warwolf.

RENAUD BEFFEYTE: Wolfwar?

____: Warwolf.

RENAUD BEFFEYTE: Warwolf. It's so difficult, we must change this name. Warwolf.

NARRATOR: It is clear from the experiment that both types of trebuchets work. Because it could so easily be increased in weight, the swinging box design was the improvement that tipped the balance in favor of attack. So the great wall-busting siege engine Edward employed at Stirling Castle was almost certainly a trebuchet with a giant swinging counterweight. The weapon that dominated siege warfare for 200 years. It was not until the late 15th century, the end of the Middle Ages, that the superiority of cannon clearly emerged, and the trebuchet vanished into the mists of time.

SPONSOR: Next week on "Secrets of Lost Empires." Up to 100 feet high, 500 tons of solid granite. The ancient Egyptians raised towers to their gods. Now modern engineers take on the challenge. Can NOVA raise Pharaoh's Obelisk?

SPONSOR: Become a medieval engineer. On NOVA's Website, build your own trebuchet and knock down the castle walls.

SPONSOR: To order the "Secrets of Lost Empires" five video set with a free family activity book, please call WGBH Boston video at 1-800-255-9424. The family activity book is also available separately, for shipping and handling only.

SPONSOR: An international tale of intrigue and suspense. The race is on to create a perfect diamond. Can science make a gem as good as the real thing? "The Diamond Deception," next time on NOVA.

SPONSOR: NOVA is a production of WGBH Boston.

SPONSOR: Major funding for NOVA is provided by the Park Foundation, dedicated to education and quality television.

SPONSOR: This program is funded in part by Northwestern Mutual Life, which has been protecting families and businesses for generations. Have you heard from the quiet company? Northwestern Mutual Life.

SPONSOR: CNET.com, helping you find the right technology products.

SPONSOR: Additional funding for this program is also provided by the David H. Koch Charitable Foundation. And by the Corporation for Public Broadcasting, and by contributions to your PBS station from viewers like you. Thank you.

SPONSOR: This is PBS.