top of page
  • Writer's pictureTastes Of History

Dispelling Some Myths: 'ancient Roman trebuchets'?

Updated: Feb 16

A recent search of the internet for information on Roman artillery machines returned, amongst other results, a website with a page entitled ‘The Roman Trebuchet Catapults’ [sic]:

We have no idea of the premise behind this website. It is very basic in design, or perhaps just in its infancy, and seems focused on being an outlet for a slideshow presentation on ‘Slavery at the Time of the Roman Empire’, authored by Garrison Hibbs and Joseph Rethemeyer. Together with the website content, the slideshow is also very simplistic, including little in the way of detail and appears aimed at a young audience. If we had to guess, it seems this is most likely an American high school project made available online. That said, if this is a first attempt to publish to the internet or is a work in progress, then that’s great. The authors are to be commended for their entrepreneurial spirit and we are only too keen to encourage the development of their future talent. With that in mind, however, if anyone openly publishes information for others to discover that is misleading, then it deserves to be challenged (as part of the peer review process). What follows, therefore, are some observations on the 'facts' cited.

The first thing to address is the page's subtitle ‘Ancient Roman Trebuchet Catapult’. For now, we can safely overlook the word ‘catapult’ as the Roman Army did indeed deploy catapultae (sing. catapulta, ‘bolt-shooters’) alongside ballistae (sing. ballista, ‘stone-throwers’) in various guises to defend towns and cities, or for use in offensive siege-works, or on the battlefield. As many as sixty of these artillery machines may have formed part of each legion’s arsenal.

No! All well and good, but the second thing to consider is the historical context and specifically what time period is being addressed. True history begins with the adoption of writing systems to record events, people, places, etc. Everything before written records is ‘pre-history’. Even so, history covers an immense span of time that only grows longer with every passing day. As a subject for study, therefore, it has become convenient to divide history into the discreet, definable periods with which we are more familiar, such as the Iron Age, the Romans, the Anglo-Saxons, the Normans, and so on. The point of drawing attention to this ‘pigeon-holing’ of history is important because the author(s) specifically used the term ‘ancient Roman’ in their title. This is important because it more precisely defines at which point in the 1,000-year history of the Romans we are focused.

The Roman empire began in 27 BC and survived into the mid-15th century AD in the (Byzantine) East, which is almost a millennium after it had collapsed in the West. The fall of Constantinople to the Ottoman sultan Mehmed the Conqueror in AD 1453 officially marks the empire’s end. Significantly, no historian today would define the 1400s as ‘antiquity’, and most would accept that the ‘ancient Roman’ world ended long before the empire’s final collapse. Yet there is still a problem agreeing when or at which point in its long decline represents the ancient empire’s physical demise.

Put simply, empires do not collapse overnight. Rome, for example, fell in a stages. The Sack of Rome by the Goths in AD 410, which saw many of the city’s finest buildings damaged, was followed by the invasion of the Vandals, who plundered the temples and took captives. The final blow came in AD 476 when a Germanic soldier, Flavius Odoacer, led a successful coup to depose the last western Roman emperor, Romulus Augustulus. So, for the purpose of history writing, and for convenience, it has become reasonable to date the end of ‘antiquity’ to the turning point year of AD 476.

No! As highlighted in the screenshot (above), the page states: ‘The exact beginning of trebuchets is not known but we do know that they first started appearing around the 12th century.’ Having just defined that ‘antiquity’ effectively ended in the 5th-century, then using the terms ‘ancient Roman’ and ‘trebuchet’ in the same sentence is an obvious, and misleading, error. So if not ‘ancient’, were trebuchets known to the later Romans of the eastern or Byzantine Empire? It seems that they were, and a lot earlier than the 12th-century.

The Strategikon is one of the most extensive extant treatises on military tactics and strategies before the early modern period and is the basis for much of what we know about the 6th-century Roman army (Olster, 2013). Significantly, it refers to a more advanced kind of artillery, recently arrived in the Mediterranean world. The text uses the classical term ‘ballista’, but this new weapon was neither a torsion nor a tension powered weapon. Rather it was one operated by traction to launch projectiles using manpower pulling on ropes at one end of a rotating beam to propel a projectile placed in a sling at the other end, thereby satisfying the Stratigikon’s description of the weapon ‘revolving at both ends’ (Dennis, 1984, 139).

What is described therefore is a traction trebuchet, also known as a ‘mangonel’ [1], pictured right. Such weapons first appeared in China in the 4th-century BC before being carried westward by the Avars. The technology was adopted by the Byzantines in the late 6th-century AD, as the Stratigikon confirms, and by their neighbours in the following centuries.

Although it required more men to operate, it was also less complex and faster to reload than the torsion-powered 'Onager' [1] that it replaced in early Medieval Europe. The author of the Strategikon does not tell us when this new kind of artillery was introduced into the Byzantine Empire, but as the treatise was written sometime after AD 580 and before AD 610 (Olster, 2013), a 6th-century date does seem likely. Moreover, the historian of Emperor Maurice's reign, Theophylaktos Simokatta, provides information on when it came into use:

‘Bousas, a Byzantine soldier captured by the Avars, taught them how to construct a siege machine, for they were ignorant of such machines. And so he prepared the helepolis to shoot missiles. With this fearsome and skilful device the Avars attacked many Byzantine cities, levelling the fortress of Appiareia in 587 and ten years later attacking Thessaloniki, which successfully resisted.’

From this description, Bousas, and other Byzantine artillerymen, must have learned how to build and operate these weapons some years before AD 587. More importantly, we learn the name the Byzantines gave the new weapon: Helepolis.

In time the traction trebuchet was replaced as the primary siege weapon by the counterweight trebuchet, also known as the counterpoise trebuchet. This later, often larger and more powerful, weapon uses a counterweight to power the arm’s swing. It appeared around the Mediterranean in the lands controlled by Christians and Muslims in the 12th-century. Ironically, given the trebuchet's origin in the Orient, the counterweight trebuchet design was carried back to China by the Mongols in the 13th-century.

And No! Finally we return to use of the term ‘catapult’. Technically trebuchets and catapults are not the same thing. How they operate can largely be differentiated by how the machine stores energy and how that energy is subsequently used to project a missile. Before the advent of gunpowder, there were three principal ways of achieving a mechanical advantage in launching a projectile.

Tension: The most familiar tension weapon is the bow. From the Stone Age onward, arrows shot from bows have been advantageous in outranging spears/javelins, throwing sticks or simple rocks to engage targets safely from distance whether in the hunt or on the battlefield. To shoot, the archer holds the bow at its centre with one hand and pulls back (draws) the arrow and the bowstring with the other (typically the dominant hand). This flexes the two limbs of the bow rearwards, which perform the function of a pair of cantilever springs to store elastic potential energy. Typically while maintaining the draw, the archer aims the shot intuitively or by sighting along the arrow. When the archer releases (looses) the draw, the limbs' stored energy is released and converted into kinetic energy transmitted via the bowstring to the arrow, propelling it forward with high velocity.

Torsion: The power of a bow is measured by its draw-weight [2]. For example, bows shot by historical re-enactors typically have draw-weights of between 50 and 70 lbs. In comparison, some of the Tudor warbows recovered from Henry VIII's flagship, the Mary Rose, have draw-weights in excess of 120 lbs making them clearly more powerful. There is, however, a limit to the draw-weight that can be pulled by the archer’s muscles alone. The introduction of torsion weapons sought to overcome this limitation, yet how and when this transition occurred remains a mystery. What is clear is that torsion - the twisting of an object by applied torque - offered much greater efficiency over tension based weaponry.

The bow was discarded in favour of ‘two wooden frames around each of which was wrapped strand after strand and layer after layer of sinew-cord’ (Marsden, 1969, 17). The two resulting bundles of sinew, each with its own frame, formed the springs. Into the middle of each of these springs was inserted a solid, tapering wooden arm. The arms were connected by a strengthened bowstring. As the bowstring is drawn rearward, typically by some form of mechanical device such as windlass, ratchet and pawl system, the arms apply torque to the sinew-cord springs overcoming inertia and thereby storing elastic potential energy. When the bowstring is released, the springs rapidly unwind to resume their resting position, accelerating the arms in a forward rotating arc. This in turn accelerates the bowstring forward imparting kinetic energy to propel the missile. Ballistas, catapults and scorpions, and other single-armed missile projectors such as the 'Onager', all use torsion power.

Traction: The earliest trebuchet, however, as described in the Strategikon, were powered by traction whereby manpower pulling on ropes at one end of a lever used the mechanical advantage of said lever to throw a projectile. Once again, the range of traction trebuchets was limited by the strength of the men pulling the ropes. Not only that but the difficulties of coordinating the pull of several men repeatedly and predictably eventually made the adoption of the counterweight trebuchet preferable even though these machines were much larger, more complicated to engineer, and thus more expensive to field. The counterweight trebuchet’s advantage, however, was using gravity to provide the throwing power.

Potential energy is stored by slowly raising an extremely heavy box (filled with stones, sand, or lead) attached to the shorter end of the lever, typically a wooden beam hinged about an axle (the fulcrum of the lever). When released, the box descends rapidly prescribing an arcing path and the force so generated causes rotational acceleration of the beam around the axle. These factors multiply the acceleration transmitted to the throwing portion of the beam and its attached sling which contains the projectile. As the beam arcs forward, the sling initially follows the same path but as the beam reaches it apogee, the sling continues to accelerate transmitting the increased speed to the projectile. The length of the sling increases the mechanical advantage, and also changes the trajectory so that, at the time of release from the sling, the projectile is traveling at the desired speed and angle to give it the range to hit the target. Adjusting the sling's release point is the primary means of fine-tuning the range, as the rest of the trebuchet's actions are difficult to adjust after construction.

The rotation speed of the throwing beam increases smoothly, starting slow but building quickly. After the projectile is released, box and the arm continue to oscillate, until the weight of the box acts as a brake to slow the rotation and bring the beam to rest, typically upright. This is unlike the violent sudden stop inherent in the action of other catapult designs such as the 'Onager' (see right), which must absorb most of the kinetic energy into its own frame, and must be heavily built and reinforced as a result. This key difference makes the trebuchet much more durable, allowing for larger and increasingly more powerful machines.

A conclusion of sorts? As stated earlier, this is not intended to be a dismissive critique of somebody else’s work. Rather, we hope it sets the record straight and explains why, in history terms at least, to connect ‘ancient’ and ‘Roman trebuchet’ is inaccurate and misleading. Moreover, it is hoped that readers can understand that while bows, catapults and trebuchets are all missile projectors, they use very different technologies to achieve the desired effect.

As ever, feel free to comment, like and share.



Dennis, G.T. (transl.), (1984), Maurice's Strategikon, Handbook of Byzantine Military Strategy, Philadelphia, p. 139.

Olster, D., (2013), ‘Strategikon of Maurice’, Wiley Online Library, Available on-line: (accessed March 27th, 2022).


1. Two common misconceptions about the mangonel are that it was a torsion siege engine and that it is synonymous with the earlier Onager.

2. A bow’s draw weight, also known as poundage, is a measurement used to determine how much force is required to flex the bow and draw its bowstring a standard length of 28 inches. Traditionally, the bow would have been held on a tiller and pound weights hung on the bowstring until it was drawn the required 28”. In this manner, a bow with a 70-pound draw weight takes 70 pounds of force to reach the standard draw length.

Recent Posts

See All


bottom of page