Non-Gunpowder Artillery

(Medieval Mechanical Artillery)

Convention is to refer to non-gunpowder artillery as 'machines' or 'engines'. While some of the the large pieces used in the late medieval era can be traced to tension and torsion powered machines used in ancient times, the rotating-beam engine (most famous of which is the trébuchet) is unique to medieval warfare. Like modern artillery, medieval crew-operated weapons were of two general categories relating to tactical employment: indirect fire (throwing or lobbing) objects (projectiles) toward a target area, and directly shooting a missile (arrow/bolt or 'bullet') at a target. The size of such weapons and logistical demands of providing the consumable projectiles to be shot emphasized the use of such machines mainly for siege operations - employed both by the besiegers and the defenders. However, such machines, in their diminutive form, were occasionally used in field battles.

Note that this page addresses only mechanical artillery and not the broader range of devices, such as towers and rams, employed in siege operations.


General Remarks
Indirect-Firing Machines:
Direct-Firing Machines Further Links
General Remarks

War machines [engines] were employed throughout the middle ages, even after the introduction of effective gunpowder artillery. The names to distinguish specific machines used by contemporary chroniclers and other observers were not standard. This leaves today's historians with some doubt, in some cases, as to what specific machine is really being described in the old texts. However, contemporary drawings and the scattered written references allow for an understanding of the various types, even though there may be some disagreement as to the exact names. Names of categories used in this page are based upon those followed by M. Reanud Beffeyte, who has researched for many years and has successfully reconstructed an array of medieval war machines. His work is further exhibited at the ARMEDIEVAL website.

Terminology is further compounded by the fact that some of the machines existed during the ancient Greek and Roman eras, and they were also used by Eastern armies, especially the Muslims. It is in the Muslim realm that the one machine unique to the medieval era emerged (most likely it had an earlier origin in Chinia) and was introduced to the Western armies. The machine, in various configurations, operated on a balance or lever principal. It propelled its projectiles from force due to the rotation of a beam about a fulcrum. These machines are often described as trébuchets, though some authorities distinguish the different configurations of these machines with distinct terms for each. Unfortunately, there is not complete agreement on the various, specific terms. This page uses the expression 'rotating-beam' machines where some other authors employ 'lever-artillery' or 'balance-machines'.

Scholars generally agree that rotating-beam throwing machines initially existed in ancient China, and are estimated to have been transmitted to Central Asia and then to the Western Muslim world around the seventh century C.E. Initially, a team of men, rather than a weight, were used to pull down the short length of a beam that rotated about a pivot point supported by an upright frame. This swiftly raised the long length of the beam which held the projectile to be launched. The use of human pulling power is described as 'traction power'. In some cases, both small counterweights and manpower were employed in these rotating-beam machines. Some authors see this combination as a 'hybrid' form in the evolution of the rotating-beam engines. In its final form, the robust rotating-beam engine operated entirely by the counterweight, and is sometimes described as a 'gravity engine'.

Stone or earth filled hutches gave the energy to launch and hurl very heavy objects. When constructed with due consideration for the geometric aspects of balance, these large machines had remarkable stability and endurance, allowing repeated firings upon the same target over a long period. Estimates are that the counterweight trébuchet was developed in the eastern Mediterranean region in the late twelfth century and its use spread rapidly to norhtern Europe. The obvious military benefit led to the counterweight rotating-beam engines replacing in the medieval era most of the ancient Greek and Roman arm-throwing catapults designs.

As with all weapons, there was continuous experimentation and variations in construction. Some rotating-beam engines had the hutches rigidly fastend to the pivoting arm. Many historians believe that these were called 'mangonels' by Western Europeans of the thirteenth through fifteenth centuries. However, the expression 'mangonel' can be traced to a Byzantine-Greek word for the rotating-beam machines used in the seventh century. Obviously the names for the machines did not remain constant as the configurations evolved during the medieval era. More on the controversy of the mangonel is addressed later in this page. In this document, the 'mangonel' will be portrayed as M. Beffeyte, a leading authority on medieval siege weapons, has judged most appropriate based upon the frequency he has found the term used by medieval observers/authors.

The most effective counterweight rotating-beam machine developed in the Western medieval armies had suspended hutches that swung on a piviot at the short end of the rotating arm. These are what M. Beffeyte calls 'trébuchets'. Expressions for 'trébuchet' ('trabuchis', 'trabuqet', etc.) reportedly first appear in Western accounts in the late twelfth century. The first references appear to be to rotating-beam machines, powered by swinging counterweight hutches, and using a sling to hold the projectile. There were variations of counterweight configurations. Some trébuchets had one hutch, others had two -- equally balanced on both sides of the beam. These were given (by some authors) a particular name, couillard.

The specific war machines shown below are divided into two broad categories according to their artillery role: Indirect Fire and Direct Fire.


Indirect-Firing Machines


trébuchet. A rotating-beam throwing engine (machine) based upon counterweight induced energy quickly rotating the arm about a pivot. The principal required balance in the structure and was most effective when using a movable hutch (à articuler le contrepoids), or hopper, that held the 'weight' of stones or earth.
Trébuchet is the more current term, though it was called many names by the medieval chroniclers and writers. Even, today there is not complete agreement on what to call the engine when the hutch is fixed, and does not swing as the arm moves (see mangonel below). The trébuchet had great precision and the advantage of hitting the same spot repeatedly. It was quite effective in trying to blast a hole through stone walls or wooden barriers, and continuing bombardment into nightime.
A form of the trébuchet which had split counterweights was a rotating-beam engine called a couillard, or biffa. The advantage was that it was lighter to transport and easier to assemble on site.
Power of rotating-beam machines varied considerably due to the weight of objects thrown and of the counterweight, among many other factors unique to each firing and engine. It is estimated that the larger machines had a range of 200 meters for hurling a 100 kg object. Its rate of fire was about 2 times per hour. Crews could be well over 50 men.




A nineteeth-century scholar's version of a traction-assisted trébuchet. It is questionable that the manpower actually added much to the effectivness of the large counterweight rotating-beam machine illustrated here. However, this 'hybrid' system was significant as the rotating-beam engines grew in size (and throwing capability) beyond the simple 'bricole' machines shown later in this page.



Controversy surrounds using the term 'mangonel' to describe a particular medieval war machine. There is no documentary support for portraying the 'mangonel' as a single-arm torsion catapult, as do many dictionaries and English/American works on medieval weapons. The convention seems to have grown from assumptions of 19th century writers who were very familiar with ancient Roman mechanical artillery, and easily conveyed the meaning of the word 'mangonel' found in early medieval sources to mean the Roman single-arm torsion catapult.
In Western documents, one of the earliest reports of a 'mangana' (Latin plural for 'mangonel') is in the poem by Abbon de Saint-Germain-des-Prés (850-923) describing the Viking siege of Paris in 885/6. However, the monk Abbon's description of the configuration of the particular engine is vague, allowing later scholars' interpretations to be based upon on their preconceived concepts of the 'mangonel'. Guillaume le Breton's report (c. 1200s) of a 'mangonellus Turcorum' also fails to provide a useful description of the weapon. The modern French translator of Guillaume, Waquet, references a French scholar on medieval weapons, Camille Enlart, who states that the mangonel was a throwing machine with a fixed, not suspended, counterweight hutch. From Enlart's work, it appears that this positon was influenced mostly by that of the 19th century French author, Viollet-le-Duc. Guillaume le Breton's reference (ca. 1200s) to a 'Turkish' mangonel as being new to the crusaders supports the theory that this was one of the medieval rotating-beam engines introduced to western Europe by the eastern Muslims.
This makes medieval Muslim documents particularly important in the search to define the medieval 'mangonel'. One such work is a well illustrated, 1462 Arabic manuscript by Yusuf ibn Urunbugha al-Zaradkash called Kitab aniq fi al- manajaniq (An Elegant Book on Trebuchets). Western Arabists examining this work and that of other Muslim writers have not yet found an unchallenged resolution to the 'mangonel' question.
A leading scholar on Arabic and early medieval literature relating to the 'rotating-beam' artillery used in the Middle Ages is Dr. Paul Chevedden. He traces the expresion 'mangonel' to the Greek term manganikón, that came into the Arabic as manjaniq. The Muslims who are credited with further developing the 'rotating-beam' engines appear to have associated the term with the larger trestle-framed configuration of this machine. Dr. Chevedden is one who practices the convention of calling all rotating-beam machines 'trebuchets'. His works are included among the references given below, and provide valuable coverage for the English reading public on the topic.

Further references:
Waquet, Henri, ed., trans. of Abbon's. Le siege de Paris par les Normands, Poeme du IXe siecle, Paris, Les Belles Lettres, 1942.
Cahen, Claude, "Un traite d'armurerie compose pour Saladin," Bulletin d'etudes orientales, 12 (1947-1948), pp.103-63.
Chevedden, Paul E. et al., "The Trebuchet," Scientific American (July 1995), pp. 66-71.
Chevedden, Paul E. "The Artillery of King James I the Conqueror." Iberia and the Mediterranean World of the Middle Ages: Essays in Honor of Robert I. Burns, edited by P.E. Chevedden, D.J. Kagay, and P.G. Padilla, Leiden: E.J. Brill, 1996, pp.47-94.
Chevedden, Paul E, "The Hybrid Trebuchet: The Halfway Step to the Counterweight Trebuchet," On the Social Origins of Medieval Institutions: Essays in Honor of Joseph F. O'Callaghan, edited by Donald J. Kagay and Theresa M. Vann, Leiden: E. J. Brill, 1998, pp.179-222.
Enlart, Camille, Manuel d'archeologie francaise depuis les temps merovingiens jusqu'a la renaissance, 3 vols., Paris, A. Picard, 1902-1916, 2nd ed., 1932, vol. 2, part 2, pp.491-92.
Hill, Donald. "Trebuchets," Viator, Medieval and Renaissance Studies, vol.4, 1973, pp.99-114.
Viollet-le-Duc, E., Dictionnaire raisonné de l'architecture française du xie au xvie siècle, vol 5, Paris, Morel, 1875.

West European and/or Muslim medieval observers may have employed the term 'mangonel' for any war machine that hurled large projectiles. The term may have had limited meaning for only a few of the medieval observers, or the meaning changed over time and among different societies. The evidence suggests that it was used most prevalently to refer to rotating-beam engines -- whether to all, or only to a particular type is unresolved. Following the count-based judgment of M. Beffeyte, this presentation classifies the mangonel as a rotating-beam throwing machine with a fixed counterweight hutch.

mangonel. Note the fixed counterweight hutch. A 'Mangonneay ..… Roues', from instructions given to the duke of Savoy's engineer in the fourteenth century.

mangonel according to Beffeyte -- many others classify it as a 'trébuchet'.

mangonel according to Beffeyte --many others classify it as a 'trébuchet'.

Note: reference to web sites with historic illustrations of trébuchets are given at the end of this webpage.



single arm, torsion-powered catapult.
The term 'catapult' is often used as a generic label for all throwing machines. In modern times it describes any system that launches an object from a platform. Military historians and reference works are not in agreement when the term 'catapult' is used to label a specifically configured medieval non-gunpowder weapon. As a specific weapon, the single arm, torsion-powered, stone-throwing engine shown here is frequently called a 'catapult'. This machine has its throwing arm's lower end inserted into twisted sinew and is held in torsion. This torsion engine is of ancinet origin. The Romans replaced the cup at the throwing end of the arm with a sling, and called the machine an 'onager'. This one-arm torsion catapult continued to be used in medieval times, though there is doubt that the term 'onager' may have still been used. Some medieval drawings suggest that some questionable modifications were attempted. One shows a large bow incorporated in the design so as to enforce the main throwing arm.
Some references label the configuration shown here as a 'mangonel' [as mentioned above under indirect-firing machines]. Many writers use the term 'catapult' to label what in this document is called a 'baliste' [see below, as another direct-firing machine].

catapult, or onager.

catapult as the Roman onager.




The term 'pierrière' was also use as a generic expression for stone-throwing engines of any kind. The general convention is to use the term for the more simple forms. They generally operated on a rotating-beam and weight-assisted [bricole] principle -- the arm rotating about a pivot supported on an upright pole-frame-- and were traction-operated by men pulling on one end of the launcing arm. No doubt, these were the first rotating-beam engines. However, they remained in the artillery inventory along with the more robust, larger 'trébuchets' (with the heavy punch) as these smaller machines were highly portable. They had the advantage of a rapid rate of fire, and remained valuable in a fire-supressing role -- against the crews of the larger artillery pieces.
Trébuchets and the small pierrières (shown here) can be classified as 'rotating-beam operated engines'. Photographs of reconstructed 'rotating-beam engines' may be viewed at a separate webpage on Siege Engines of the Middle Ages.


Direct-Firing Machines


tension-powered arbalète à tour.

two arms, torsion-powered baliste [ballista].
The large, carriage or post-mounted crossbow had two distinct forms, even in ancient times. One configuration, arbalète à tour, was of a single bow, aligned horizontally; its energy was from tension (as with the normal hand-bow), due to the spanning of the wood or steel bow. The second configuration, baliste [ballista], was two horizontally-aligned arms, each under torsion when the common chord at the far ends was spanned. The spanning of either device required wenches [arbalète à tour]. Each of the configurations could be modified to project sizable stones as well as large arrows or javelin bolts. These were often portable enough to be brought on to some battlefields, but were more commonly used in position defense. There is strong doubt that the two arm, torsion-powered configuration continued to be used in medieval times. Most accounts suggest that the tension bows, firing bolts or arrows were what the medieval forces used as their principal heavy direct-firing weapon until the advent of the gunpowder pot-de-fer in the early 14th century.

Several references label the large, post-mounted crossbow a 'catapult', and call what was shown earlier in this document as a 'catapult', a 'ballista'. This convention appears to be a carry-over of those used with the study of ancient military history, where the Roman 'ballista' [a 'catapult/onager' in this document] had two smaller configurations termed 'onager' and 'scorpion'. The 'scorpion' of the medieval era [see below] was not the same as the one used by the Romans. A few reference works label the medieval large, mounted crossbow a 'Springal'.


spring-engine [a scorpion shown here].
As the concept was simple, there were many variations of spring-engines. It was characterized by the main launching, or striking, arm being bent back and held under tension. When the arm was released, it struck and launched an arrow (like the image shown), it was a true direct-firing weapon. Some variations had the bent-back arm be a 'throwing-arm', topped with a sling or cup, from which stones and the like were hurled. These had a slight curved-trajectory, and performed as indirect-fire weapons, much like the catapults shown earlier.

RETURN To Top of Page.


To see modern reconstructions of some of the engines on this page, visit ARMEDIEVAL website. The site reflects a few of the over 30 medieval mechnical engines, and some early gunpowder weapons, constructed by M. Renaud Beffeyte and used in sustained demonstrations at various locations.

Another interesting aspect to explore is The Secrets of the Medieval Engineer.

Medieval Siege Engine Kits for educational instruction are promoted at the following website:
Pathfinders Design and Technology: Medieval Siege Engine

Kits are promoted as easy to make, fun and educational. The model machines accurately recreate medieval siege weaponry, and are not so big that they will hurt someone. The kits are by Derek Wulff, an inventor, teacher and wooden science kit creator. He teaches classes on building workable bridges and simple machines with the goal to enhance an understanding of medieval times, as well to encourage children to make and use toys that celebrate ingenuity and science.

More information on medieval artillery can be found at
Gunpowder Weapons of the Late Fifteenth Century website.

Return to Hundred Years' War Web Page.
Return to Oriflamme Web Page.

This page was created in July 1998, revised and last updated 28 May 2004.
Comments can be sent to the Société de l'Oriflamme.