The Queen’s University of Belfast (2003)
While there has been much debate over when the traction trebuchet appeared in the west and the extent to which it replaced late-antique torsion weapons there, its appearance in eastern Europe and the Mediterranean has attracted less attention. This cannot afford to be passed over without comment, especially given its implications for later military development, as the trebuchet, in its various forms, was the mainstay of medieval siege artillery until the development of effective cannon. Some earlier historians such as Hurri reckoned that the trebuchet came to western Europe when the first crusaders returned home at the close of the eleventh century. However this view was challenged in the 1970s when Donald Hill asserted that the route of transmission was almost certainly through the Arabs, a theory supported by Carroll Gillmor. This view ignores evidence in the seventh-century Miracles of St Demetrioswhich brings the Byzantines and Avars into the picture, as recognised by James Howard-Johnson and Paul Chevedden. Sometime in the late sixth or early seventh century this form of lever artillery began to appear in the Byzantine empire, probably spreading from China. Were the Avars, a nomadic people from the steppes, the means of its dissemination to the Byzantines, and, if so, how quickly did the Byzantines, and possibly the Persians, copy this weapon? The introduction of the traction trebuchet is a very difficult issue to resolve owing to a lack of clear source material and the question is still open to much debate. This paper does not set out to draw any definitive conclusions; rather its purpose is to reopen discussion on the topic through the introduction of new elements to the equation.
The adoption of the early traction trebuchet is an interesting example of technology transfer. Our main source for its introduction comes from the Miracles of St Demetrioswhere they describe the Avar siege of Thessalonica. The bishop who penned the miracles, John, compares the inaccurate rope-pulled devices of the Avars with the accuracy of the Byzantine torsion artillery. This is most interesting as it shows that the Avar weapons were noticeably different from the Byzantine ones. The bishop gives the following description:
‘….they were four-sided; they rose from broader bases to narrower tops on which there were massive cylinders, their ends sheathed with a thick layer of iron; to them there were attached lengths of wood, like the beams of a large house, and these had slings; and when these (the slings) were raised up, they sent out rocks; the rocks were large, the shots frequent…’
This is clearly a description of a trebuchet and compares well with pictures in medieval manuscripts, apart from the fact that the men pulling on ropes, the propellant of the traction trebuchet, seem at first to have escaped the bishop’s observation. However, they may well have been obscured from his view by the sheaths of iron, or probably more commonly leather, which could have been used as protection for the otherwise exposed crew members. It is likely that the trebuchet was introduced from the east around this point in time, and the Avars could well have been the means of its transmission from China where it had been in service since the third century BC.
Based on the evidence in the Miracles of St Demetrios it can be assumed that the traction trebuchet made its debut in the latter half of the sixth century or possibly the early seventh, depending on the date of the siege and when the account was written by Bishop John. Certainly in the second half of the seventh century it became more common at the expense of torsion artillery. As most armies naturally tried to adopt weaponry superior to their own, it is not surprising that the Byzantines would have attempted to copy and improve on the trebuchet as used by the Avars, but when might that have happened?
This particular question brings us to what may be a new piece of the trebuchet jigsaw, the siege of an unnamed fort near Akbas attacked by the Byzantine general Droctulfus in the 587. Theophylact points out that a lot of missiles missed their target, flying over the fort. Since the Persian counter devices were impact-reducing mats, the missiles in question are probably large stones, as opposed to ballista bolts. Could these overs be due to crews being unfamiliar with new weapons? This imprecision is in stark contrast with the usual remarks about the accuracy of Byzantine artillery, and echoes the inaccuracy of the Avar machines recorded by Bishop John. Could this be an indication of the Byzantines starting to use and operate their own trebuchets? There is no concrete evidence to support this. The Byzantine crews could simply have been bad shots or may have been working with poorly constructed weapons. However the remark about their inaccuracy is in stark contrast to all other comments about Byzantine artillery and the dating of the siege ties in with the possible appearance of this machine in Europe.
This coincidence is intriguing, but the most critical thing is that Theophylact says that the missed shots were overs, in other words the range of the machines was greater than was needed. Had the Byzantines placed their artillery too far up a hill overlooking the fort? The area is very mountainous, but the odds are against the offensive side actually climbing above the fort and dragging their artillery up with them, either in parts or completed, probably because the fort itself, still unlocated, would have been on the highest suitable spot already. If the machine was not more powerful than usual then the only other explanation is that the projectiles were lighter and travelled further than expected. However the trebuchet is generally believed to have had a shorter range than standard torsion artillery, with a trajectory more akin to a mortar than a cannon. Could a trebuchet have fired overs? We will probably never know, but while the standard medieval trebuchet may have had a shorter range, it is interesting to note that Bishop John says some of the Avar shots in the siege of Thessalonica were too short and others too far. Did early machines appear to be more powerful simply because they threw lighter missiles?
The question of when this form of artillery was adopted, and why, is more difficult. The two pieces of evidence which apparently counter the early 580s as the date of transmission come from two different sources. Theophylact, describing the siege of Appiaria in 587, says that this was the first occasion on which the Avars used siege machinery, and a Byzantine deserter, Busas, passed on this knowledge. The Avars obviously had the ability to take towns before this and the deserter story could be complete fiction. Apart from this difficulty there is the fact that the story is not precise in its detail. The weapon in question may be a siege tower or an artillery piece. Theophylact Simocatta refers to Busas telling the Avars how to construct a siege machine. ‘kaˆ dÁta Bous©j to›j ‘Ab£rouj ™d…daske sump»gnusqai poliorkhtikÒn ti mhc£nhma, œti tîn toioÚtwn Ñrg£nwn ¢maqest£touj Øp®rcontaj ¢krobol…zein te pareskeÚaze tæn ˜löpolin.’ The key word is ¢krobol…zein since if it means ‘from on high’, then the description may refer to a siege tower which overtopped the walls, but if it means ‘from afar’ then we are probably dealing with an artillery piece. If the story is true and artillery is what is inferred, it would be referring to the standard Byzantine torsion variety. This machine would have been complex and difficult to build for unskilled engineers. Does it imply that the Avars had no artillery of their own at this point? Certainly none is mentioned by Theophylact before this event. Would the Avars have been able to copy and operate these torsions weapons sufficiently accurately? Would they have wanted to have torsion engines?
Despite Theophylact’s silence it is probable that the Avars did already have stone throwing weapons, assuming the date of 586 to be true for the siege of Thessalonica, and if they had not been shown them by the Byzantines, they must have had their own. If the account of an Avar siege of Thessalonica in the Miracles of St Demetrios is to be believed, we are told they pulled on ropes on their machines. This suggests lever artillery, not the standard Byzantine torsion engines. The fact that the Avars had captured several cities in 586 further supports, though it can by no means prove, the view that they had artillery since it may have helped them to accomplish this task. If it is torsion artillery that was betrayed to the Avars, they may have been interested in it because of its superior accuracy and range compared to the lever weapons that they themselves had been using. If this was indeed the case then the Byzantines may have encountered the trebuchet before 587, since these were surely the weapons the Avars intended to complement or replace. If the Avars did not have any artillery before this date, where did they suddenly find out about the trebuchet before their attack on Thessalonica? Therefore the value of this story about Busas is very dubious; it simply raises additional questions to which there are no answers.
This brings us to the other item of counter evidence, the 14th Miracle of St Demetrios itself, which concerns the second siege of Thessalonica. The date of this event itself is open to much debate which further confuses the matter, but while some scholars put it at 597, I prefer the date of 586, as advocated by Lemerle. Nevertheless, even if 586 is correct, and is the first instance of the Avars using the trebuchet against the Byzantines then is it likely that Droctulfus’ men were discharging similar devices a year later? Had the Byzantines come across the device earlier? As noted above this is possible. Could they have had access to a trebuchet before their contact with the Avars? Had the Persians adopted it first as a result of clashes with the nomads and the Byzantines copied it from their Sassanian foe? Did the empire come across it when dealing with another nomadic group?
In addition to the issue of when the device was first seen and copied by the Byzantines, there is the problem of why they would copy it. Bishop John is very clear in his contrast of the accuracy of the two types of artillery. Why would the Byzantines copy something that was less accurate than the equipment they already possessed? The Avars had not been able to approach Diocletianopolis because the city’s mural artillery was good enough to keep them away from the walls. Therefore if the bishop’s observation about the superiority of the Byzantine equipment was probably true, why should the empire copy inferior enemy weapons?
The faster rate of fire which the trebuchet is also credited with by Bishop John is unlikely to have been the main reason for its adoption by the Byzantines since there was little merit is wasting ammunition on inaccurate shots, unless the misses were near enough to make the defenders keep their heads down. However the Byzantines may have realised that it required trained teams of pullers to make the machines into accurate catapults, and possibly better construction technique, particularly around the fulcrum. Had the Byzantines realised the greater power of the trebuchet which would potentially be of use in trying to batter down defences, a task which their standard torsion artillery does not really appear to have been used for? The possible increase in power that could come from building larger versions of the machine may have been appreciated. There are few examples of very large torsion weapons, probably owing to the size of the springs and frames, and they are only mentioned in a defensive capacity. Larger defensive weapons mounted on walls may have created too much recoil when fired. Was there really any use for such large mural pieces anyway, when smaller items seem to have been capable of beating back enemy siege engines? On campaign, they may have been too awkward to take apart and transport. This was a gap the trebuchet could fill. Powerful shots which led to collateral damage were almost certainly the main advantages that the trebuchet possessed over the standard Byzantine artillery of the time and that is why it was copied.
Another possible benefit of the trebuchet may have been the simplicity of its construction. Keeping stores of springs in good order, especially in damp weather, maintaining the weapons and training gunners were inherent problems associated with torsion weapons, both ballistae and onagers. However, the trebuchet was a simple device. In its most basic form it was an entirely wooden machine with one pivot, also wooden, as the only moving part. Men had just to be trained to pull in unison. Such weapons would surely have been easier to produce in quantity than any torsion artillery pieces.
However, while the Byzantines are probably the first users of the trebuchet in Europe and Asia Minor, it cannot be assumed that they had a monopoly on it. The Avars are not mentioned as attackers of the Persian empire, but they may have been in contact with the Sassanian forces. There is every possibility that they attacked towns north of the Persian frontier and could have employed the trebuchet there. Details may have passed to the Persians. More importantly, if the weapons deployed by the Byzantines near Akbas were copies of Avar trebuchets, the Persians would have seen them. Lacking artillery of their own in any quantity, this device would have been the object of much interest. It would have been apparent that it did not require springs or metal. Since metal was scare in Persian lands and arguably may have affected their output of military weapons, a simple wooden stone-thrower capable of hurling large rocks would have been a desirable item. Its simplicity would ensure its rapid construction and distribution. While there is no clear evidence of Sassanian employment of this device, their siege of Jerusalem in 614 does raise one possibility. The Persians breached the walls of the city with ballistae according to Antiochus. While ballistae had been capable of firing bolts or small stones in their earlier days, the stone-throwing role appears to have disappeared by early Byzantine times. Clearly these weapons were not used to batter walls down, but if the word ballistae is being used to mean artillery then this is the first case we have of it causing a breach, rams or tunnels being the usual means of putting holes in circuit walls. If artillery had done this, it is likely that there had been a major advance in Persian weaponry by the early seventh century. Had the Persians built large trebuchets? This is a very tenuous argument and corroborating evidence on the siege from Sebeos states that the walls were brought down by mining operations.
Irrespective of their possible employment by the Sassanians, there is now no doubt that the Arabs did indeed have trebuchets in the seventh century. The word used by the Arab historians to describe their catapults is mandjanik, which comes from the Byzantine manganikon, itself derived from the early Greek manganon meaning a device, particularly lifting devices like cranes, which also used levers and fulcrums. Manganon was also used to cover siege engines, but the more technical term manganikon emerged in the seventh century.The word is first used in the Chronicon Paschale describing the Avar machines at the siege of Constantinople in 626. This itself is an early work further reinforcing the belief that the Avars were the people who introduced the trebuchet to the Byzantines. The fact that the Arab word used to describe their catapults, probably trebuchets, is derived from Greek suggests that they learnt of the weapon from their western imperial neighbours, rather than the Persians. The western term usually taken to indicate trebuchets is mangonel, although Bradbury has shown the danger of trying to be dogmatic about weapon types based purely on the word given to them in western sources. The fact that in the eighth century the Arabs adopted arradah from the Syriac arada meaning wild ass, or onager in Latin implies that they also learnt about torsion artillery from the Byzantines as well, another argument against the disappearance of these particular weapons, in the east at least, and that they drew a distinction between different types of catapult.
In addition to the eastern dimension of the trebuchet issue, the Avars are also known to have been in contact with the west. They had several campaigns involving the Franks and Lombards and we know that some Slavs were present at most if not all of the Avar sieges. In 603 Slavs were present at the successful sieges of Cremona and Mantua in Italy as allies of the Lombards, who had requested help from the Avar Khagan in their struggle with the Byzantines, under the terms of the everlasting alliance concluded between these two parties and the Franks. In the latter attack at least, battering rams were employed, their first recorded use by the Lombards. It is tempting to believe they were brought by the Slav contingent. While there is no way of proving this, we do have one example of this weapon being used in an Avar siege, and it is likely their Slav vassals knew about battering rams, if it was not they who actually built and operated them. Was it just a coincidence that the Slavs were present the first time battering rams were used, or did they bring the weapons, or at least the means of construction, with them, possibly in return for the Lombards sending shipwrights to the Khagan? Unfortunately it is impossible to say for sure, but it is a big coincidence. However this opens up the possibility of the Avars, perhaps through the Slavs, being the channel for the trebuchet into western Europe as well as Byzantium.
The purpose of most artillery weapons before this date must also be remembered. Contrary to popular opinion they do not appear to have been used to destroy walls. Rather the majority of descriptions referring to artillery suggest that it was primarily an anti-personnel weapon, and archery was just as efficient in that respect, though it lacked the punch of heavier torsion machines. This was certainly the case on offence, though some of the heavier stones discharged may have been able to dislodge crenellations and batter gates open. Defensively artillery was still used against attacking infantry and cavalry, but the larger mural weapons were also used to combat the attacker’s siege machines. Joshua the Stylite mentions a machine, the crusher, which lobbed an enormous stone of over 300lbs. While it would not have been impossible to construct a machine capable of hurling such a weight, Joshua may be exaggerating. However, in his defence, the machine seems to have been employed against enemy engines and sheds near the wall, so the large stones would not have to travel very far. Other mentions of giant stones include Ammianus’ description of the loss of the siege towers at Aquileia in 361, the siege of Theodosiopolis in 421/2, and the Arab siege of Silos in 663. There is no reason to believe that varying sizes of onager were not in use. This could explain the huge stones which are mentioned in some siege accounts. There were certainly different ballistae sizes so why should there not be variations in onagers as well? Since none of them seems to have been small items and they would consequently have been difficult to bring along or construct in situ during a campaign, it is further support for the use of artillery in a primarily anti-personnel role.
However the introduction of the trebuchet meant that relatively simple weapons capable of discharging ammunition heavy enough to batter through walls could be constructed in situ during a siege. This was its real potential value. It is not clear when this was realised, nor is it clear when these weapons entered service, but the siege of Synnada in 663witnessed the deployment of heavy weapons by both the Byzantines and Arabs. It is just possible, though on balance unlikely, that Jerusalem was the first Byzantine victim of such Persian built equipment in 614? There were earlier cases of heavy artillery pieces, such as Joshua the Stylite’s Amidan ‘crusher’. Whitby believes that heavier weapons capable of throwing loads over greater distances than before may have been the reason for some changes in town circuits, notably at Antioch.
The descriptions of Byzantine siege machines, particularly artillery, are increasingly vague in the fifth and sixth centuries. Could lever artillery possibly have been introduced during this period? The Byzantine weapons mentioned by Procopius at the siege of Rome in 537 include ballistae, though possibly with an iron stave as opposed to two sinew springs, and the stone throwing onagers. After this date the word onager is not used, but the machine did remain in use despite the introduction of the trebuchet. Without clear evidence to back up the less clear descriptions, the Avars must remain clear favourites as the first users of the trebuchet in the Byzantine world. It is possible that the Avars had used this artillery on other towns before Thessalonica, since they had captured quite a few in the preceding years. There are too many possible interpretations of the source material to say that the Byzantines were definitely using trebuchets by 587 or even when it was first used in the region, but any debate on the introduction of this machine must weigh all these arguments very carefully. However, although it is unlikely that the Byzantines were using a device they had seen only a year earlier in quantity, it is certainly not impossible that they could have attempted to copy such a relatively simple machine in that time span, and the weapon was probably in use by the late 580s.
Whatever the reason for the adoption of the trebuchet by combatants from the seventh century onwards, the weapons still do not seem to have made as great an impact as might be expected, at least not immediately. Writing at the close of the tenth century, Nikephoros Ouranos says, ‘The men of old, in their pursuit of siege warfare, constructed many devices such as battering rams, wooden towers, scaling ladders with various features, tortoises, and all kinds of other things which our generation can hardly imagine. It has, however, tried all these devices and found that out of all of them, the most effective way, one the enemy cannot match, is undermining the foundations, …..’
K.Huuri, Zur Geschicte des Mittelaltreichen Geschützwesens (Helsinki, 1941). L.White Jnr, Medieval Military Technology and Social Change (Oxford, 1962).
D.Hill believed that the trebuchet was first positively mentioned in Arab use in Al-Tabar ’s account of the siege of Mecca in 683, ‘Trebuchets’, Viator, 4 (1973), 99-114. C.M.Gillmor, ‘The Introduction of the Traction trebuchet in to the Latin West’, Viator, 12 (1981), 1-8
Miracula Sancti Demetrii, ed. and tr. P.Lemerle, Les plus anciens recueils des miracles de Saint Démétrius et de la pénetration des Slaves dans les Balkans, vol I (Paris, 1979). J.Howard-Johnson, ‘Thema’, Maistor ed. A.Moffatt (Canberra, 1984), 189-197. P.Chevedden, ‘The Artillery Revolution of the Middle-Ages: The Impact of the Trebuchet on the Development of Fortifications’, unpublished paper, 1992 appears to accept Byzantine involvement, however I have been unable to gain access to this article and only know of it through K.Devries, Medieval Military Technology (New York, 1992), 134. J.Howard-Johnson was kind enough to send me a copy of an excellent, though unpublished, article of his, ‘Byzantine Artillery’, drafted in 1971, dealing with precisely the issue of when the trebuchet appeared.
According to K.Hurri the first evidence of the trebuchet in Byzantine service comes with Al-Bundar ’s description of a tall stone-thrower operated by 1200 men in 1071. K.Huuri,Geschützwesens, 91. C.Gillmor, ‘Introduction’, 1-2, n4, seems to accept this in her article concerning the appearance of the machine in the west. She goes on to argue for a ninth-century date for the introduction of the trebuchet, but her evidence, based on the descriptions of stone-throwing machines, is not decisive. Not one is as clear as that of the catapult recorded in the Miracula Sancti Demetrii. A more realistic view of the debate over the introduction of the trebuchet and the silence in western sources concerning torsion weapons in the early middle ages is put forward by J.Bradbury, The Medieval Siege (Woodbridge, 1992), 250-270.
The date of the transmission of the trebuchet from China to western Europe has been debated over the last 50 years. Some believe it appeared in the west when the soldiers of the First Crusade encountered it in use by the Turks. However, most modern historians accept that it was certainly known to the Arabs from an earlier date. Discussion with Dr James Howard-Johnson when he visited Belfast in 1994 provided the spur to investigate the matter further, and I am indebted to him for his help and assistance in the early stages of my research on this topic. S.McCotter, The Strategy and Tactics of Siege Warfare in the Early Byzantine Period: From Constantine to Heraclius (Unpublished PhD. The Queen’s University Belfast, 1995), 128 and 212-213.
A traction trebuchet is one where power is imparted to the transverse beam by men pulling ropes attached to its shorter end, the beam being pivoted off centre. A counter-weight trebuchet relies on a heavily weighted box replacing these men to provide the downward impetus at the short end which will catapult the projectile at the opposite end. The counter-weight device is more consistent than a human powered device.
Mirac.St Demet, XIV.153.
There is good reason to believe that the source of this information, the author of the Miracles of St Demetrius, is describing the operation of machines from his own time and simply assuming that weapons worked in the same way half a century earlier, as pointed out by Whitby. L.M. Whitby, The Emperor Maurice and his Historian, Theophylact Simocatta on Persian and Balkan Warfare (Oxford, 1988), 118.
Dr Howard-Johnston’s translation. For an alternative translation see K.Devries, Technology, 134.
J.Needham, ‘China’s Trebuchets, Manned and Counterweighted’, On Pre-Modern Technology and Science: Studies in Honour of Lynn White Jnr, eds B.S.Hall and D.C.West (Malibu, 1976), 107-45, notices the similarity of design between Chinese trebuchets and those depicted in the twelfth-century manuscript of Pietro di Eboli’s Carmen de Rebus Siculis.
We know that the Byzantines were prepared to adopt other military equipment from the Avars such as their cloak and possibly the stirrup. The composite bow was used by the steppe nomads and the imperial forces appreciated its qualities sufficiently to copy it, particularly after the fifth century. Nomads had produced other novel forms of siege equipment such as the Hun mobile elevated firing platform used at Naissus and the Sabir Hun light battering ram. The latter was quickly employed by both settled powers. Any nomad force with designs on settled regions, mainly the areas to the north of the Persian and Byzantine realms, but often northern areas of these empires themselves, needed to have some poliorcetic capabilities. As they were coming from the east and probably had contact with the Chinese and their advanced civilisation, it should not be surprising that they had some novel pieces of siege equipment and other weaponry. Either as a result of being in friendly contact with them, serving in their armies or by attacking their towns, the nomads could have acquired practical knowledge of these devices. The trebuchet is one example of siege machinery that came west and there is no good reason to doubt that the Avars may have brought it and the Byzantines copied it.
Theophylact Simocatta, Historiae, II.18,3-4. Ed. I Bekker (CSHB, 1834); tr. L.M. Whitby, The History of Theophylact Simocatta (Oxford, 1986).
Mirac.St Demet, XIV.154.
This story about Busas is far from convincing, though many have taken it at face value, like K.Devries, Technology, 134. There are many cases in Byzantine literature of deserters and traitors helping enemies to get into towns or passing information to them. The whole story of Busas being captured, the Byzantines refusing to pay his ransom money and his subsequent treachery is very suspicious. Theoph.Sim., II.16,1-11. Nevertheless it is possible that it is true, or at least there is a basis of truth in that some captured imperial engineers may have passed the information on either willingly or under duress.
They had taken eight towns the previous year.
The use of torsion artillery by the ‘barbarians’ is rejected by many historians for different reasons. Some believe that it was too complicated for them while others reckon that they had little use for it any way.
Mirac.St Demet, XIV.154.
S.Vryonis, ‘The Evolution of Slavic Society and the Slavic Invasions of Greece: The First Major Slavic Attack on Thessalonica, AD 597’, Hesperia, 50 (1981), 378-90.
Miracula Sancti Demetrii, ed. and tr. P.Lemerle, Les plus anciens recueils des miracles de Saint Démétrius et de la pénetration des Slaves dans les Balkans, vol II (Paris, 1979), 150-153. The arguments for 586 are nicely summarised by Whitby in The Emperor Maurice, 117-121.
Atilla is reputed to have had artillery at the siege of Aquileia in 452, artillery that was capable of breaching walls, a feat not normally credited to torsion weapons. However the sources are not conclusive on this point. Priscus, f.22; Procop., Wars, III.4,30-35; Prosp., 1365; Marcell.Com., s.a.452; Greg.Tur., HF, II.7 and Gall.Chron.511, 617. Priscus, referring to the siege of Aquileia, says all kinds of artillery were brought up. Unfortunately, even this fragment is suspect, since ‘armamentorum’ had been seen as another possible reading for ‘tormentorum’ in one edition, and ‘machines’ may simply be the most appropriate translation. Priscus, f.22.1 ‘…qui machinis constructis omniaque genera tormetorum adhibita,…’. (Jord., Get, 222). R.C. Blockley, The Fragmentary Classicising Historians of the Later Roman Empire, II (ARCA, 10, Liverpool, 1983), 310. If it was constructed on the spot, as the text of Priscus implies, then that makes it less likely to have been artillery, at least torsion artillery, and therefore battering rams may be the most likely weapons to have been built. The Huns could have had artillery of some sort by this date, even if it was only captured equipment. Without further information in the sources, it is pointless speculating on the Hunnic firepower, but it does leave open the possibility thay they, another steppe grouping, were the earliest users of lever artillery.
There is no source evidence to suggest that this actually happened, but there is generally very little information on Persia and the nomads on her northern borders from this period. We know that the Huns made some forays into their territory and I see no good reason to believe that the Avars were not capable of doing so either.
In 337 the Persians supposedly did not have enough iron to equip their army. Lib., Or, LIX.66f. The eastern frontier merchants were not allowed to export bronze and iron to the Persians. Exposito, 22.
The Persians’ lack of defensive artillery in accounts of the period is initially a source of surprise, but the aforementioned scarcity of metal may explain this. However, it should be seen in the context of relatively few mentions of artillery in sources, compared to the number of sieges examined.
Ant.Strat, 506. ‘They shot with such violence from their ballistae that on the twenty-first day they broke down the city wall.’
 The Armenian History attributed to Sebeos, trans. R.W. Thompson and J.Howard-Johnson (Liverpool, 1999), 69.
J.Howard-Johnson, ‘Byzantine Artillery’, 7.
Chronicon Paschale, ed. L. Dindorf (CSHB, 1832).
Bradbury, Siege, 252-254.
Paulus, HL, IV.28.
Agilulf had sent some shipwrights, probably Byzantines, to the Avar Khagan in 601. Perhaps the two had reached some sort of technology exchange arrangement. Paulus, HL, IV.20.
S. McCotter, Siege Warfare, 120-139 for artillery and missile weapons in late antiquity.
The ballista was the most commonly mentioned artillery piece in late antiquity, and since it generally fired bolts rather than stones, it was an anti-personnel weapon (though it was sometimes fitted with incendiary ammunition). The onager and stone throwing devices were less frequently mentioned.
Theod., HE, V.36.
A Byzantine master carpenter agreed to build a catapult big enough to take the fortress for the Arabs if his family were spared. The first shot hit the gate but failed to break it, while successive shots fell ever shorter. The Byzantines used their own catapult to return fire which shot such a huge stone that it smashed the Arabs’ weapon and killed several men as it rolled away. Maronite Chron., AG.975.
Maronite Chron., AG.975.
L.M. Whitby, ‘Procopius and Antioch’, The Eastern Frontier of the Roman Empire, eds D.H. French and C.S. Lightfoot, (BAR.Int.Series, S553, Oxford, 1989), 541.
Procop., Wars, V.19,14-19. However, the actual term onager had only ever been used previously by Ammianus Marcellinus. The general lack of technical precision in siege descriptions where the most technical term used is normally manganon/mechanis or machinis makes it difficult to say how frequently each type of weapon was ever used.
Taktika, 65.22-23. Tr. E.McGeer, ‘The Taktika of Nicephoros Ouranos’, DOP, 45 (1991),138.