Proceedings 31       18 February 1993        the Warburg Institute, Woburn Square, London

THE TREBUCHET: ITS HISTORY AND DYNAMIC: a talk by DONALD HILL (member SHMTS) whose important article "Trebuchet" was published in the University of California Medieval and Renaissance Studies VIATOR Vol.4 (1973).

The President reminded members of Mr Hill's international reputation for his work on Islamic technology, before inviting him to take the floor.


There were two main types of trebuchet:

[a] the traction machine, comprising a beam rotating about an axle with bearings support. on a timber tower sometimes wheeled. At the long arm of the beam was a sling carrying the missile; at the short end a fitting to which ropes were attached. The machine was operated by a team of men pulling on these ropes.

[b] the counterweight machine likewise comprised a pivoted beam, but it was operated by the power of gravity, actuated by a counterweight suspended at the end of the short arm of the beam. The beam was spanned by a winch, which was released by a special hook.


From the Qin dynasty of China (249-202 BC) we find TRACTION TREBUCHETS commonly mentioned in accounts of sieges, though the first detailed description dates from AD 759. The major work on the subject was a treatise of 1044 which detailed methods of construction as well as the performances of various types. The device was encountered by the advancing Arab armies in southwest Asia in the late 7th century and a reference to a stone-throwing weapon (manjaniq) called Umm Farwa (The mother of hair) at the siege of Mecca in 683 may allude to the dangling ropes. The word manjaniq, meaning as it does simply any stone throwing device, must be treated warily but the following seem to be references to the traction trebuchet: 708, 'The Bride', with a Moslem army in Sind, worked by a team of men under a skilled operator who aimed and 'shot' the device; al-Fahja (the 'straddlelegged’), 710; and the manjaniq at the siege of Baghdad in 865, worked by men who pulled on ropes to let fly the missiles.

A description of a siege engine found in the late 10th century encyclopaedia, Mafatih al-Ulum ('Keys of the Sciences') of al-Khuwarazmi named the parts, e.g.: the tower support as the 'chair', the fulcrum as 'the sow', and the beam as 'the arrow' (sahm). Although there was no reference to ropes or sling this could be presumed to be a TRACTION TREBUCHET. Khuwarazmi's terms 'the sow' and 'the arrow' were also used by Marda b. Ali from his military manual for Saladin (ruled 1169-93), which contains the first description of counterweight trebuchet. Arab armies continued to use the traction machine even after the introduction of the counterweight device. Judging from the number of such weapons recorded in sieges of 13th century crusader castles, they may even have preferred it, It discharged lighter missiles but was easier to build, lighter and therefore more mobile and had a greater rate of fire. Although there were some 12th century references to the traction trebuchet in Europe, e.g. the northern crusaders' siege of Muslim Lisbon (1147) it never matched the popularity of the counterweight device there.

Marda b. Ali called the COUNTERWEIGHT MACHINE manjaniq farsi, i.e. a 'Persian trebuchet', but writers' nomenclatures were too diverse to set great store by this. First recorded at the siege of Acre (1189-91) when it was deployed by both Crusaders and Moslems, the time and place of invention are unknown. However specialist artificers soon appeared the trade of 'trebuchator' being known in England, for example, in the early 1200s.

Construction and Performance

The most detailed accounts of TRACTION TREBUCHETs came from China's Sung dynasty (960-1280). We find two basic types of support [a] a single timber post with the beam swivel at the top. The post could be rotated inside a substructure which facilitated aiming. This was the lighter weapon with comparatively restricted range. [b] a more robust timber frame support of trapezoidal cross section up to 8 metres high and 5 tonnes weight. Here the beam, (>8.4 metres long) comprised several spars (> 15) bound together. The number could be changed to suit battle conditions, but elasticity had to be retained

At the narrower end (up to 7 cm diam.) was a copper nest, attached to the spar by iron wire, thus forming a sling. At the wider end (up to 12.5 cm) was the special fitting, apparently a crossbar fitted with large rings, to which were attached the ropes, up to 125 in number, 15.5 metres long and 16 mm thick. As many as 250 men might be assigned to a machine, possibly two men to a rope on larger ones. The beam was divided into two arms by the axle in the ratio of long to short 5:1 or 6:1 for light machines or 2:1 or 3:1 for larger ones.

The range of Sung machines varied from 75 to 150 metres and the weight of the missiles, usually spherical stones, from 1 to 60 kg. They were apparently built to standard specifications and from the same materials. Such standardization favoured rapid replacement of damaged parts in battle and production of huge numbers of weapons, 800 in the year 979 alone. According to Persian sources, the army of Genghiz Khan had a Chinese artillery corps of some 10,000 men.

Arabic accounts are similar in broad outline, though Marda b. Ali describes machines of only single spars. He lists Arab, Turkish, Byzantine and Frankish models all basically similar with maximum range of 120 yards. Apparently light machines which achieved their best results with a flexible beam made of cherry wood. A vital member of the artillery team was the rami or "shooter". He held the pouch containing the missile against his chest with all his force; it was important that he did so at the correct angle, otherwise there would be no discharge, or the stone would fly vertically upwards. Presumably he let go of the pouch the instant before the men pulled hard on the ropes. (but see illustration 2! Ed.)

Despite many references and ms illustrations there was surprisingly ...