Strategies of Defence, Problems of Security: the Garrisons of Constantinople in the Middle Byzantine Period

Map_of_Constantinople_(1422)_by_Florentine_cartographer_Cristoforo_BuondelmonteStrategies of Defence, Problems of Security: the Garrisons of Constantinople in the Middle Byzantine Period

John Haldon

Constantinople and its Hinterland: Papers from the Twenty-Seventh Spring Symposium of Byzantine Studies, Oxford, April 1993 (1995)

The title I have chosen is intended to convey a particular paradox which confronts the defenders of any capital city, and Constantinople in the middle Byzantine period – which I will take, rather generously, to mean the period from the 6th to the 12th century – is no exception to this. This paradox resides in the simple fact that, as capitals, seats of government or imperial households, centres of social attraction and wealth, and so forth, such cities have invariably come to hold this position because they are well situated in respect of communications, accessibility, transport and, of course, supplies and provisions.[1] Constantinople would not have become an imperial centre had this not been the case. I am sure we can all think of the occasional exception to this, but in principle it holds true. On the other hand, however, this very accessibility makes such centres attractive to aggressors or invaders, both because they are perceived as foci of political and ideological power, and because they can actually be reached, overland or by sea, without too much difficulty. They are big, and they are attractive targets.

In consequence, the rulers or governments of states centred at such important focal points may be expected to take certain more-or-less predictable measures to deal with this situation. I want here to look at the ways in which the late Roman and Byzantine state organised the defence of its capital, and at what other questions defending a city like Constantinople inevitably raises.[2]


Let us first of all, then, briefly review the ways in which the city was garrisoned and defended over the period in question. Defending an urban centre such as Constantinople is achieved in two main ways: by passive means, and by active means. The former involves walls, earthworks, ditches, towers and so on – ways of blocking a hostile force with physical obstacles. The latter involves soldiers and the movement of soldiers, materials, provisions and so on, to the end that the defenders have an advantage in each of these respects in contrast with the attackers. At the beginning of our period, which is to say the period up to the midle of the seventh century, we find the passive defence of the city depended essentially on two factors: the great Theodosian land walls and their extension in the Blachernai region[3], and the attached sea-defences[4], supplemented by the so-called Long Walls in Thrace (which as we have heard had only a limited effectiveness), within the region behind which a certain number of troops were based[5]; and the fact that the city was a peninsula, with only one side accessible by land forces. In addition, of course, the crucial factor of water-supply, and therefore of the construction, maintenance and repair of aqueducts, cisterns and related works; and of foodstuffs (in particular grain, and the related issue of its import, distriibution and storage) must be taken into account in any full treatment. The last two issus will not be dealt with in this paper.[6] Its active defence rested on the presence of two field armies, under the magistri militum praesentales or in praesenti, based in Bithynia and Thrace[7]. There were in theory, although we have only very approximate figures, about 20-25,000 soldiers under each commander; but in practice, many of these troops were attached to other field armies during war-time; while a proportion of them were always unfit for active service due to illness, injury, malingering and so on[8]. Nevertheless, these are sizable forces, and blocked the way to any attack on the city itself, especially from the more exposed northern and western fronts. In addition, there were also present in and around the City, independent of the two magistri praesentales, a number of palatine units, of various categories. In the sixth and much of the seventh century, the most important of these were the excubitores, created during the reign of the emperor Leo I, a small elite bodyguard of 300 picked soldiers, originally mostly from Isauria.[9] But this was an imperial bodyguard, rather than an element of any formal defensive force for Constantinople. In addition, there were seven units of scholae, originally crack cavalry units, each of 500 or so men, recruited mostly from barbarian, especially German, sources in the fourth and fifth centuries, but by the sixth century they had become a parade-ground elite dominated by young men with contacts at court and the appropriate private rsources to purchase their commissions. Individual scholares, officers and men, did fight on active service, usually seconded from their units; and larger groups may similarly have seen campaign service. But for the most part the evidence is that they fulfilled a formal, ceremonial job, and were not much value from the military point of view. They were based mostly in Bithynia and Galatia, in cities such as Prusa, Nicaea, Nicomedia, Dorylaion and so on, but one schola always appears to have been in Constantinople.[10] Apart from these, there were other, much smaller groups of palatine soldiers, but again they hardly counted towards the defence of the city itself. The only group deputed specifically to the defence of the city appears to have been the Watch, under the prefect of the city, referred to as the pedatoura or kerketon in the seventh century and after, which doubled up as an urban police force to keep the peace and enforce the city’s laws. Its size is unknown, but it appears not to have been very numerous.[11]

Throughout the seventh and into the eighth century, at least until the reign of Constantine V, the imperial government seems to have relied on the regional field armies for the real defence of the city. On several occasions during this period, however, these troops were absent on campaign, and the defences of the city were entrusted to those actually inside it. These seem always to have been relatively few in number, and recognition of this is apparent in the measures adopted to man the Theodosian and other defences, in particular the sea-walls, in times of crisis. Especially important for this were the Constantinopolitan corporations or ‘guilds’ – each corporation, whether butchers, jewellers or whatever, had a duty to man a particular stretch of wall. The demes of Blues and Greens played a central role in this, and although it must be stressed that they did not constitute a permanent urban militia or anything of the sort, they were clearly an important part of the city’s defence – weapons were issued to them, as to the other corporations, when the government deemed it necessary, and this is actually recommended for cities in general in the so-called Strategikon of Maurice.[12] Thus in 559, 601, 602 and 610, the corporations, including the Blue and Green demes, as well as the scholae in the city and others such as the excubitores were ordered to defend the Theodosian walls or otherwise engage in military duties;[13] indeed, in 559 members of the senate were also ordered to assist in the defence against a threatened Avar attack (although this probably refers to the retinues of senators, where they possessed them); while the scholae and the city’s civic defenders, the corporations, went out to meet the attackers near the Long Walls, where they were driven back, perhaps not surprisingly.[14] There is less information about the situation during the middle and later seventh century, but what there is confirms the impression that the actual garrison of the city was always minimal, with the Watch under the prefect’s authority constituting the main military or para-military force, in control of the gates and towers of the walls, as well as internal security. Proper defence thus depended entirely upon the regional field armies in Thrace and Asia Minor. When these failed, then emergency measures were taken. In exceptional circumstances, as we shall hear shortly for 626, for example, then extra troops, in quite large numbers, could be drafted in to form a proper defense capable of both beating off a sustained siege and attack and sallying out to harry and harass the enemy army.[15] Similarly in both 674-678 and 717-718 field troops from the praesental and other armies appear to have been drafted in to strengthen the defence of the City.[16]

The beginnings of a change in this strategy can be detected in the later seventh century. In the context of his apparently very unpopular policies, Justinian II fortified, or improved the defences of, the palatine region, and seems to have established some new units to man these defences, although the evidence is uncertain. At some point, the baths of Zeuxippos partly served to house these troops, with the result that the buildng eventually became known as the noumera. Later, one of these units was responsible for the Chalke gate of the palace, so that between them they provided a garrison for the imperial palace precinct (the domestic of the noumera and the komês tôn teicheôn commanded these forces, the latter having nothing to do with either the Theodosian walls of the City or the Long Walls in Thrace, as the texts make quite clear).[17] But the important point about this development is that it marks the beginnings of the establishment of new regiments in Constantinople which were intended to defend the emperors and the government against a threat from the city and the other soldiers who might be based there.

During the later seventh century, through the reign of Leo III and into that of Constantine V, the threat posed to the emperors by the possible disaffection of the main field forces, upon which the defence of Constantonople ultimately rested, forced the emperors to introduce certain changes in an effort to improve their security as well as that of their city. Leo III tried to do this by giving key positions to trusted officers and relatives, most importantly his son-in-law Artavasdos, who became commander of the Opsikion, the field force garrisoned in Bithynia and neighbouring districts of Asia minor, and from which the garrison troops for CP appear normally to have been drawn.. But Artavasdos’ rebellion in 741/2 showed that this policy provided no long-term solution; and Constantine V eventually embarked upon a different policy. In the first place, he seems to have been responsible for breaking up the old Opsikion into a number of smaller regions whose soldiers and commanders posed thereby less of a threat[18]; in the second place, he also reformed and reorganised the palatine units of the scholai and the exkoubitores, the latter having also, during the later seventh and early eighth century, become litle more than a ceremonial body. Of course, this last move should also be understood in the context of Constantine’s iconoclast politics, which meant that he needed a powerful and loyal force at his disposal within the city both to protect him and to impose his policies effectively. But there is evidence that the re-organisation of these units, henceforth referred to in the sources as the imperial tagmata to distinguish them from the themata, took place before they were turned into a powerful pro-iconoclastic tool.[19]

The numbers of the tagmata remain disputed, although they do not appear to have been more than a few thousand at the outside.[20] They were both an imperial elite guard and an instrument of imperial politics; but they certainly also served as an element in the field army, and the sources often refer to them as forming the core of imperial expeditionary forces. But while they could also function as a much more powerful and more-or-less permanent garrison, this was only as long as the emperor was present. There is a good deal of evidence to show that the tagmata were not conceived of as a garrison for the city – they normally accompanied the emperor on campaign, indeed, they could be sent away even when the emperor did not go to war. And although successive emperors created new tagmata or drafted in provincial units to secure their own political position, the evidence again is that they normally accompanied the emperors if they were away. Thus Eirene drafted in the regiments that eventually became theVigla, Nikephoros I created the Hikanatoi and brought in the Phoideratoi from Lykaonia, Michael II recruited the so-called tessarakontarion and the Hetaireai, and so on. How many of them remained in Constantinople when the emperors were present is unknown, a point to which I will return in a moment. After their creation, the units known as the noumera and the Walls do seem to have been permanently based in CP, but they were there to defend the palace (although there is no reason to think that, in a crisis, they would not also be deployed for the defence of the city).[21]

There is thus no evidence that Constantinople possessed a large, permanent garrison. If there was a known threat, provincial and full-time troops were brought in to defend CP; where no threat was apparent, the urban militia and the few guards attached to the Prefect of the city, together with any residual imperial guards and ceremonial units, seem to have been seen as sufficient. A ninth-century text makes this quite clear: when the emperor left Constantinople, he appointed a representative to take charge of affairs. Under this person’s authority was the prefect of the city, with his urban Watch, the kerketon, and the corporations or systêmata; together with the commanders of the noumera and Walls regiments. In all the sieges of or attacks upon the City for which we have evidence after the later eighth century: 821-822 (Thomas the Slav vs. Michael II), 860, 907 (the ‘Rus), 913 (Symeon of Bulgaria), 988 (Bardas Phgocas vs. Basil II) and so on – there is no suggestion that any other troops than these, and sometimes units of the imperial tagmata which happened to be present, were involved, unless brought into the City by the emperors (which happened anyway in most of the examples listed above). The Arab historian al-Tabarî even notes thst the emperor Leo VI persuaded Muslim prisoners to help defend the City during a Bulgar attack in 896-7.[22] It is worth noting that this occurred immediately after a major Byzantine defeat at Bulgarophygon, so that the capital was no longer covered by any regular field forces.[23] And if we pursue the story through the eleventh and twelfth centuries it will be found to be no different. During the rebellion of George Maniakes in 1043, the elite units of Varangian and Norman troops were with the rebel, while at Constantinople Constantine IX possessed only the prefect’s Watch and some palatine ceremonial units. In 1047 during the rebellion of Leo Tornikios, similarly, the imperial capital had no large garrison which could face the attack. In 1081, it was the commander of a unit of the regular (mercenary) troops – mostly German, Norman and Varangian-English mercenaries – who eventually let Alexios I enter.[24] Of a regular Constantinopolitan garrison, apart from Botaneiates’ own troops, there is no word. Indeed, it is clear that the citizens of Constantinople, who played such an important role in imperial politics (at least in respect of their influence when emperors pursued what to them were unpopular policies), positively objected to the presence of large numbers of soldiers in their City – in 641/2, when the general Valentinus used his provincial forces to impose himself on the young Constans II, as well as under Nikephoros II in the 960s, whose Armenian soldiers were deeply unpopular with the populace.[25]

It may seem strange that the capital city of such an empire should be so lightly defended. But we need to bear a number of factors in mind. In the first place, the strength of the defences was considerable. Even in the siege of 1453, the defenders of the City, who had to man both the whole length of the Theodosian walls as well as considerable stretches of the sea-walls, numbered only about 8,000, of whom only about 5,000 were Byzantines, and of these most were not soldiers. Yet this small force was able to repel the first two waves of the Ottoman assault on May 29th, an attacking force of probably 20-30,000 men if not more; and the Janissaries themselves succeeded in gaining entrance only after an hour of their general assault had been launched, when fifty or so (out of the probable 5,000) forced the small Kerkoporta postern near the Kaligaria gate (mod. Egri Kapi). When it is borne in mind that, according to the available evidence, it was thought sufficient against such numbers to place 200 or so Genoese crossbowmen for the defence of the whole mile-long length of wall from the Golden Gate to the Pêgê or Selymbria Gate, the effectiveness of the land defences can be appreciated. The heaviest concentration of Byzantine and foreign soldiers was along the stretch of walls from around the Gate of St. Romanos to the Charisios or Adrianople Gate, against which the main attack was directed; even here, less than a thousand soldiers held back the first two waves and the full strength of the Janissaries before they were outflanked as a result of the fall of the Kerkoporta.[26] When it is further borne in mind that the Ottoman forces were greatly helped by more than a month of massive artillery bombardment, which seriously damaged the walls (which were anyway too weak to support Byzantine artillery firing back), the point I am trying to make should be clear. Providing the gates were secured and the defences provided with a skeleton force, the City was safe against even very large forces in the pre-gunpowder period which concerns us.

In the second place, the cost of maintaining a permanent and sizeable garrison, in view of the relative security of the city’s defences, was considerable. Serious threats to the imperial capital were actually not that frequent; when there was a threat, regional forces could usually be got to the city quite quickly, even in the critical situation of the 626 siege. The cost of maintaining its armies was the single largest item of imperial expenditure; the army was the main consumer of fiscal revenues during the sixth century, and there is no reason to think that this changed thereafter. Maintaining large numbers of troops permanently at Constantinople would have been particularly problematic.[27] During the seventh and eighth centuries, the state saved direct expenditure by promoting the provincialisation and localisation of the armies, so that large elements became, in effect, a self-supporting militia. But full-time soldiers still had to be maintained, whether in the provinces or near the capital, and their costs also had to be covered.[28] The tagmata especially cost a good deal to support; and a balance had to be maintained between military effectiveness and numbers, and expense. And military costs went up again in the tenth and eleventh century as an ever larger proportion of the imperial armies was maintained on a full-time basis and paid directly by the treasury.[29] The figures we possess all suggest that the central elite units in the tenth century numbered perhaps 4,000 at the outside; but we should not underestimate the effectiveness of 4000 heavily armed cavalrymen. Basil II turned the tide of a civil war with 6000 Varangians, and Alexios I thought that 500 Flemish knights would be sufficient to give him a substantial advantage against the Turks in Asia Minor. On the whole, armies in the medieval period were really quite small compared with the preceding and following eras. There are good historical reasons for this, which we need not go into here. But Byzantine armies are no exception.[30]

In the third place, large numbers of troops in or around the capital, however loyal in theory, always represented a threat to the emperor’s safety. Hence the small numbers in imperial guards regiments such as the excubitoresin the fifth and sixth century (300), and the preference for soldiers of foreign, i.e. non-partisan, origin in such units. The reliance on regular field armies for the defence of the City, as opposed to a permanent city garrison, seems on the whole to have worked. When it failed, as in the later seventh and first half of the eighth centuries, it was because the political context, the nature of the relationship between the praesental armies (the Opsikion) and the capital, the character of the Opsikion army itself (in respect of its sources of recruitment and relationship to the other field armies) had all changed; while the dependence of the capital on the Opsikion in the context of the Arab and Bulgar wars had increased very considerably. Constantine V redressed the balance by creating a centrally-controlled field force, the tagmata, which served as a counter-weight to these provincial forces. But the tagmata themselves then became a potential threat, and so each emperor had to re-establish the balance with their own chosen units.

Defending the capital, therefore, involved a constant balancing act between financial, political and logistical factors. And in the light of these, it seems not surprising that no permanent military force of any size was ever stationed permanently within the city walls. Those units which were in Constantinople were kept, to a degree at least, under separate chains of command, to minimise the possibility of their uniting against the emperor of the day. But this brings us to my next set of questions: how were soldiers housed in the capital, and where?

Archaeological evidence for the sites of barracks is virtually non-existent. The bottom rooms of the large towers on the inner Theodosian wall were certainly used by soldiers, but whether as a permanent barracks or merely on those occasions when extra troops were drafted in is unclear.[31] Perhaps this was where the prefect’s watch were billeted when on duty. Some accommodation in the palatine precinct had been provided by the sixth century for the candidati and excubitores, as well as for the scholae, as evidenced by the names of various related structures in the palace area; but while this might have housed the first two groups, it could not have housed more than a fraction of the latter, 3,500 strong.[32] Here, the evidence is clear, however: the scholae served in CP on a rotational basis, so that only one unit, or part of several units, were present at a time. The later noumera and the Walls regiment were probably housed in the structure later known as ta noumera, as mentioned already, a part of the Zeuxippos baths complex;[33] the tagmata established by Constantine V seem at first to have been largely based within the city walls, for textual references, although not corroborated by any independent tradition, suggest that Constantine converted various churches and monastic buildings, in particular the monastery of Dalmatos, for use as stables and barracks for his guards.[34] Some evidence to support this, or at least to support some sort of secularisation which might fit in with this, comes from archaeological sources, for the nature of the numismatic evidence in the Saraçhane (St. Polyeuktos) excavations suggest a possible secular use for parts of the church during the eighth and first half of the ninth centuries.[35] But Eirene replaced the tagmata recruited by Constantine V with her own guards unit, and certainly by the later ninth century, they were billeted outside, in Macedonia, Thrace and Bithynia.[36] An inscription of about 843 records the construction of a barracks of some sort, referred to as the ‘arsenal’, under Theophilus and Michael III.[37] The Vigla appears to have been responsible for the hippodrome area and the adjacent palatine precinct, and may also have been quartered there, although this is uncertain.[38]

It must also be remembered that there were considerable open areas between the Theodosian walls and the old walls of Constantine, in which soldiers might have been based at times. In addition, it is known that field troops sometimes encamped outside the Theodosian walls, and there were also military encampments at the Hebdomon, a few miles away from the City itself.[39] The solution to the problem of the district called tafla or talaya in ninth- and tenth-century Arab geographers’ accounts of the Byzantine provinces, a district situated between Constantinople and the Long Walls, appears to lie here – the Arabic talaya seems to be a rendering of the Greek ta allagia. And allagion, pl. allagia, was a standard term for a military unit (derived from the changing or rotating of watches or guard-duties).[40]

As well as the defences and soldiers of the capital itself, of course, there were also further lines of defence which might serve to hinder or obstruct an enemy advance on Constantinople. In Thrace the military base of the Hebdomon, seven Roman miles from the City itself, was the site of a parade ground and major military encampments, and the small towns or fortresses of Herakleia, Selymbria (at which point the Long Walls of Thrace commence) and Rhegion covered the approaches along the extension of the via Egnatia; Arkadioupolis and Bizye covered the main route from the North; and in Asia Minor Malagina, Nicaea and Nicomedeia covered the approaches to the City from the South and East.[41] All these were defended fortress-towns in Byzantine times (although Malagina may have been established as a base only in the later eighth century[42]), and occur frequently in the written sources as places where the Byzantines confronted invaders, or each other in civil wars. If they failed to halt a hostile attack, they enabled the defences of the capital to be brought in order, allowed a breathing space to permit the entry of troops into the City. Accounts of imperial responses to Arab or Bulgar attacks make this role explicit, and the care with which the central government seems to have spent on the fortifications at Malagina and Nicaea, for example, illustrates this function. Malagina had other functions, too, serving as an imperial aplêkton or military camp, and an imperial stables and base for the imperial expeditionary baggage-train.[43] In addition, of course, during the ninth century, and possibly earlier, there had been the system of beacons linkung the eastern frontier with the capital.[44]

This brief survey of the ways in which the imperial government defended its capital has brought out two main points. First, the issue of security was vital to the emperors, with the result that large concentrations of soldiers in Constantinople seem generally to have ben avoided. Certainly no institutional arrangements existed for the maintenance of large bodies of soldiers there on a permanent basis, except briefly during the reigns of Constantine V and Leo IV, and possibly under Nikephoros I. Second, the cost of maintaining such forces, and the formidable problems associated with supplying and housing them in the City, must also have deterred emperors from keeping big garrisons in the City. Such a policy also gave a heightened importance to those fortresses and strongpoints covering the main routes to Constantinople, and explains the importance they held. In the end, therefore, the defence of the capital city of the empire depended upon the maintenance of a careful balance between fiscal considerations and questions of imperial security. Given the limited resources available to most emperors, it seems that it was, in the end, financial concerns that tended to dominate.

End Notes

[1] See the remarks made by J. Koder in respect of settlement-patterns and resource-distribution: ‘The Urban Character of the Early Byzantine Empire: Some reflections on a Settlement Geographical Approach to the Topic’, in The Seventeenth International Byzantine Congress, Major Papers (New York 1986) 155-187.

[2] Most general works on Byzantine history have brief comments on the city’s defences and its garrisons, but for the most part discussion of the former is confined to descriptions of the walls, and discussion of the latter is vitiated by an inadequate knowledge of the structure, administration, financing and numbers of the military units in question. The most recent general discussion of the whole subject is B.C.P. Tsangadas, The Fortifications and Defense of Constantinople . East European Monographs no. LXXI (New York 1980) which includes also a good bibliography of the standard works up to the date of publication.

[3] Tsangadas, op.cit., 7-32; esp. G. Meyer, A.-M. Schneider, Die Landmauer von Konstantinopel. 2 vols. (Denkmäler antiker Architektur VI & VIII. Berlin 1943); A.-M. Schneider, Die landmauer von Konstantinopel. Sitzungsberichte der Preussischen Akad, der Wisenschaften, phil.-hist. Kl. 32 (Berlin 1933) 1157-1172; R. Janin, Constantinople byzantine. Développement urbain et répertoire topographique (Paris 21964)248ff. with maps; and on the evolution of the plans for the walls, their realisation, and the Constantinopolitan legends that developed around them, P. Speck,’Der Mauerbau in 60 Tagen: Zum Datum der Errichtung der Landmauer von Konstantinopel mit einem Anhang über der Notitia urbis Constantinopolitanae’, in Studien zur Frühgeschichte Konstantinopels, ed. H.-G. Beck (Miscellanea Byzantina Monacensia 14. Munich 1973) 135-178; and on the Blachernai walls, Meyer- Schneider, vol. 2, 95-122; P. Speck, Zufälliges zum Bellum Avaricum des Georgios Pisides (Misc. Byz. Monacensia 24. Munich 1980) with modern literature.

[4] See Tsangadas, op.cit., 33-59 for a convenient summary; A.-M. Schneider, ‘Mauern und Tore am Goldenen Horn zu Konstantinopel’, Nachrichten d. Akad. d. Wiss. in Göttingen, phil.-hist. Kl. (1950) Nr. 5, 65-107; also Janin, Constantinople byzantine, 218ff, 269ff.; and the remarks by C. Mango, ‘Recent Turkish Works on the City’s Walls’, Speculum 30 (1955) 271f.

[5] See M. Harrison, ‘The Long Walls in Thrace’, Archaeologia Aeliana, 4th ser., 47 (1969) 33ff.; B. Croke, ‘The Date of the “Anastasian Long Wall” in Thrace’, GRBS 23 (1982) 59-78; A.H.M. Jones, The Later Roman Empire 284-602: A Social, Economic and Administrative Survey (Oxford 1964) 656. The most recent discussion of the question of the troops based in the area enclosed by the long walls: J.F. Haldon, Byzantine Praetorians: An Administrative, Institutional and Social Survey of the Opsikion and Tagmata, c.580-900 (Poikila Byzantina 3. Bonn 1984) 271ff.

[6] The most useful brief modern account of these issues is to be found in C. Mango, Le développement urbain de Constantinople (IVe-VIIe siècles) (Paris 1985); see also the relevant entries (Constantinople; Aqueduct etc.) in The Oxford Dictionary of Byzantium, 23 vols., ed. A. Kazhdan et al. (Oxford 1991).

[7] See Jones, Later Roman Empire, 609, 655; Notitia Dignitatum Utriusque Imperii, ed. O. Seeck (Leipzig 1876), Or. v, vi.

[8] For the approximate numbers, see J.F. Haldon, Byzantium in the Seventh Century. The Transformation of a Culture (Cambridge 1990/revised edn 1997), 251-253; and Jones, Later Roman Empire, 679-686.

[9] See Haldon, Byzantine Praetorians, 136-139, 161-164.

[10] Jones, Later Roman Empire, 613-614, 657; R.I. Frank, Scholae Palatinae. The Palace Guards of the Later Roman Empire (Papers and Monographs of the American Academy in Rome, 23. Rome 1969); Haldon, Byzantine Praetorians, 119-128; 142-161 for detailed discussion and sources.

[11] The Prefect’s soldiers were under a kentyriôn in the ninth and tenth century (Kletorologion of Philotheos, in: N. Oikonomidès, Les listes de préséance byzantines des IXe et Xe siècles (Paris 1972) 81-235, see 113.14, and comm. p. 320), who may or may not be the same official as the biglomagistôr mentioned in a seventh-century text (R. Devréesse, ‘Le texte grec de l’Hypomnesticum de Théodore Spoudée’, AB 53 [1935] 66-80 [BHG 2261] 72.24); they were referred to as the kerketon, vigla (not to be confused with the imperial tagma of the same name) or pedatoura, reflecting both their patrols through the City and their duties as watchmen on the walls and gates of the City (Epanagoge Basilii, Leonis et Alexandri, in: JGR [Zepos] ii, 229-368 [cf. A. Schminck, Studien zu mittelbyzantinischen Rechtsbüchern {Forschungen zur byzantinischen Rechtsgeschichte 13, Frankfurt a. M. 1986} 1-15], see iv.8). On the prefect’s rsponsibilities in this respect, see also G. Dagron, Naissance d’une capitale. Constantinople et ses institutions de 330 à 451 (Paris 1974) 277ff. See Haldon, Byzantine Praetorians, 541-542, note 769 for further references; and idem, Constantine Porphyrogenitus, Three Treatises on Imperial Military Expeditions. Introd., edition, transl. and comm. (CFHB XXVIII. Vienna 1990), comm. to (C) 310 and (C) 420 (pp. 232, 240) for the military content of the terms.

[12] I have discussed the evidence for this in full in Byzantine Praetorians, 258ff. For the reference in the Strategikon of Maurice, see Das Strategikon des Maurikios, ed. G.T. Dennis, trans. E. Gamillscheg (CFHB 17, Vienna 1981) x, 3.32-35 (Eng. trans. G.T. Dennis, Maurice’s Strategikon. Handbook of Byzantine Military Strategy [Philadelphia 1984] 109); and see A. Cameron, Circus Factions. Blues and Greens at Rome and Constantinople (Oxford 1976) 107ff., 120ff.; Dagron, Naissance, 356f..

[13] 559: Theophanes (Theophanis Chronographia, ed. C. de Boor, 2 vols.[Leipzig 1883, 1885] 233.12-22 (and cf. Agathias [Agathiae Myrinaei Historiarum libri V, ed. R. Keydell{Berlin 1967} v, 16.2); 601: Theophanes, 287.21seq.; 602: Theoph. Sim. (Theophylacti Simocattae Historia, ed. C. de Boor [Leipzig 1887; revised and emended edn. P. Wirth, Stuttgart 1972]) viii. 7-8; 610: Joh. Antioch. (in Excerpta historica iussu imp. Constantini Porphyrogeniti confecta III: Excerpta de Insidiis, ed. C. de Boor [Berlin 1905] 58-150) 150 (also ed. as Ioannis Antiocheni Fragmenta, in: Fragmenta Historicorum Graecorum, edd. C. and Th. Müller, 5 vols. [Paris 1874-1885] v, 27-38, see 38 (frg. 218[f]);

[14] Theophanes, 233.16-18; Agathias, v, 14.

[15] There appears to have been only a small body of soldiers left at Constantinople (see Chronicon Paschale, ed. L. Dindorf [CSHB, Bonn 1832] 717.5seq.) until Heraclius divided his field forces into three and reinforced the capital (Theophanes, 315.11seq.; Theodore Sykellos (Peri tês tôn atheôn barbarôn kai Persôn … kinêseôs kai anachôrêseôs, ed. L. Sternbach, in: Analecta Avarica [Rozprawy Akademii Umiejetnosci Wydzial Filologisczny, ser. 2, vol. 15. Cracow 1900] 2-24 [= L. Sternbach, Studia Philologica in Georgium Pisidam {Cracow 1900} 297-334]) 302.8seq.; G. Pisid., Bell. Av. (George of Pisidia, Bellum Avaricum, in A. Pertusi, ed., Giorgio di Pisidia. Poemi 1. Panegirici Epici [Studia Patristica et Byzantina 7, Ettal 1959] 176-200) 280. According to the Chron. Pasch., 718.18-22, there were some 12,000 cavalry soldiers in CP in 626, some of whom were Armenians (ibid. 724.11; cf. G. Pisid., Bell. Av. 280f.). These must have been mostly those sent by Heraclius. There is no report of their arrival, but they are described as elite troops (Chron. Pasch., loc.cit.); while Heraclius’ brother Theodore arrived later, as the siege was raised (Chron. Pasch. 726.4-10), having encountered and defeated the army of the Persian general Sahin in the East (cf. Theophanes, 315.17-22). The reinforcements sent by the emperor to CP had presumably arrived long before this. F. Barisic, ‘Le siège de Constantinople par les Avares en 626’, B 24 (1954) 371-395, 378, dates their arrival to the time between the despatch of Athanasius as envoy to the Khagan at the end of June or early July, which fits in well with the events that followed (although his estimate of the number of soldiers in the capital at this time [‘quelques douzaines de milliers’ – ‘Le siège’, 391, note 3] seems very exaggerated). A. Pertusi, ‘La formation des thèmes byzantins’, in: Berichte zum XI. Internatioonalen Byzantinisten-Kongreß I (Munich 1958) 1-40, see 25, thinks that these 12,000 were part of Heraclius’ praesental army left behind; this seems unlikely, since he needed all the troops he could muster in the East, while he hoped that he had made a truce with the Avars (Theophanes, 302.27-30), and the very fact that he felt obliged to send a considerable force back to Constantinople militates against this view. Neither do these soldiers appear to have formed the corps supposedly stationed along the Long Walls of Thrace (pace Speck, Bellum Avaricum, 44-45 and note 215): as I have shown elsewhere, such a force never existed (Byzantine Praetorians, 271-275). There remains the problem that Theodore arrived in the region opposite the City, so that Bonus was able to point to the army and threaten the Khagan. Yet at the same time Sahrbaraz and his army were supposedly still at Chalcedon (cf. Theophanes, 316.25-27; Theodore Syncellus, 313.14-27). No conflict is recorded. Speck noted the difficulty (Bellum Avaricum, 44-47), but has conflated the relief army despatched by Heraclius, with that under Theodore’s command, chiefly on the grounds that the three-fold division of Heraclius’ army resembles a literary-theological topos, and because the arrival of this force is not mentioned But neither is the arrival of that under Theodore, Heraclius’ brother, mentioned by either Theodore Syncellus or George of Pisidia (Bellum Avaricum, 45 and notes). While the fact that the account of the three-fold division of the army was attributed with a symbolic importance by Pisides does not by itself mean that it did not happen. It seems more likely that the first relief army was able to gain entrance to the City before the main siege on the landward side of the city set in, transported by the ships obviously available to Bonos, in command of the City during the emperor’s absence (see Theodore Syncellus, 311; Chron. Pasch., 724; Nikephoros Patr., Breviarium [ed. De Boor], 18.6-24 [ed. Mango, ]); while Theodore arrived, as Speck suggests, in time to discourage the Persians under Sahrbaraz encamped on the Asiatic shore of the Bosphoros and force their withdrawal, as Barisic originally suggested (and see K. Ericsson, ‘Revising a Date in the Chronicon Paschale’, JÖBG 17 (1968) 17-28, 24; A. Stratos, ‘The Avars’ Attack on Byzantium in the Year 626′, BF 2 (1967) 370-376, see 372..

[16] See the accounts of Tsangadas, op.cit., 107ff., 134ff.; and R.-J. Lilie, Die byzantinische Reaktion auf die Ausbreitung der Araber (Miscellanea Byzantina Monacensia 22, Munich 1976) 77ff., 16ff., who provides a detailed account of the context and development of the sieges.

[17] For full discussion, see Haldon, Byzantine Praetorians, 256ff. On the Zeuxippos, see C. Mango, The Brazen House: a Study of the vestibule of the Imperial Palace of Constantinople (Arkeologisk-Kunsthistoriske Meddelelsev udgivet af der Kongelige Danske Videnskabernes Selskab, Bind 4, nr. 4, Copenhagen 1959) 28, n.27; 39-42; Janin, Constantinople byzantine, 169-170.

[18] See H. Gelzer, Die Genesis der byzantinischen Themenverfassung. Abhandlungend. königl. sächsichen Gesellschaft d. Wiss., phil.-hist. Kl. (Leipzig 1899/Amsterdam 1966) 79, 91; Ch. Diehl, ‘L’origine du régime des Thèmes dans l’empire byzantin’, in: idem, Études Byzantines (Paris 1905), 276-292; G. Ostrogorsky, History of the Byzantine State (Oxford 1968) 140. See the commentary of Lilie, Die byzantinische Reaktion, 323ff. for a good summary of the strategic role of the Opsikion at this time.

[19] Haldon, Byzantine Praetorians, 205ff., 228-235.

[20] For the alternative views, see W. Treadgold, ‘Notes on the Numbers and organisation of the Ninth-Century Byzantine Army’, GRBS 21 (1980) 269-288 (arguing for high numbers); Haldon, Byzantine Praetorians, 276-282, 629-633; F. Winkelmann, ‘Probleme der Informationen des al-Garmi über die byzantinischen Provinzen’, BS 43 (1982) 18-29 (arguing for low numbers).

[21] On all these points, see Haldon, Byzantine Praetorians, 245-256 with sources and literature.

[22] See A.A. Vasiliev, Byzance et les Arabes, II, éd. fr. H. Grégoire, M. Canard (Corpus Bruxellense Hist. Byz. II, Bruxelles 1968) ii, 11-12.

[23] See D. Obolensky, The Byzantine Commonwealth. Eastern Europe, 500-1453 (London 1971) 106.

[24] See M. Angold, The Byzantine Empire 1025-1204: A Political History (London 1984) 37ff., 50, 104.

[25] Valentinus: Theophanes, 343.3-5; Nicephorus, Breviarium, [de Boor 29.19seq.; Mango, ]; Sebeos (F. Macler, trans., Sébéos, Histoire d’Héraclius (Paris 1904) 103, 105-106; Nicephorus II: Leo Diaconus (leonis Diaconi Caloensis Historiae Libri Decem, ed. C.B. Hase [CSHB, Bonn 1828])iv. 6-7 (63.14-65.1).

[26] Good accounts can be found in S. Runciman, The Fall of Constantinople, 1453 (Cambridge 1965) 91-139; and, with especial detail of the troop dispositions, M. Bartusis, The Late Byzantine Army. Arms and Society, 1204-1453 (Philadelphia 1992) 123ff.

[27] For the salaries and costs of the tagmata in the tenth century, for example, see the figures at haldon, Byzantine Praetorians, 308; and for the costs of armies in general M.F. Hendy, Studies in the Byzantine Monetary Economy c.300-1453 (Cambridge 1985) 158-159, 221ff. The figures given by W.T. Treadgold, The Byzantine State Finances in the Eighth and Ninth Centuries (New York 1982)12ff., 69ff., although not entirely to be dismissed, are nevertheless vitiated by the tendentious nature of the relevant sources. See esp. the critique by R.-J. Lilie, ‘Die byzantinischen Staatsfinanzen im 8. und 9. Jahrhundert und die stratiôtika ktêmata’, BS 48 (1987) 49-55. See also the recent survey by C. Morrisson, ‘Monnaie et finances dans l’empire byzantin: Xe-XIVe siècle’, in Hommes et richesses dans l’empire byzantin, II: VIIIe-XVe siècle (Paris, 1991), 291-315, see 297 (although following uncritically the figurs offered by Treadgold, op.cit., and unaware of the points raised by Lilie, art.cit.).

[28] See R.-J. Lilie, ‘Die zweihundertjährige reform: zu den Anfängen der Themenorganisation im 7. und 8. Jahrhundert’, BS 45 (1984) 27-39, 190-201; Haldon, Byzantium in the Seventh Century, ; idem, ‘Military Service, Military Lands and the Status of Soldiers: Current Problems and Interpretations’, DOP 47 (1993).

[29] See Morrisson, ‘Monnaie et finances’ (cited above, n. 26); N. Oikonomidès, ‘L’Évolution de l’organisation administrative de l’empire byzantin au XIe siècle’, TM 6 (1976) 125-152.

[30] See my remarks in Byzantium in the Seventh Century, , with literature and sources.

[31] Cod. Theod. (Theodosiani libri xvi cum constitutionibus Sirmondianis, edd. Th. Mommsen, P. Meyer et al. [Berlin 1905]) vii, 8.13

[32] R. Guilland, Études de topographie de Constantinople byzantine I, II (Berliner Byzantinistische Arbeiten 37. Berlin 1969) I, 3-40: ‘Les quartiers militaires’.

[33] See the references at note 16 above; and Guilland, op.cit. I, 41-55: ‘Les Noumera’

[34] Theophanes, 443.1-5; note also J.D. Mansi, ed., Sacrorum Conciliorum nova et amplissima Collectio (Florence 1759-1927) 329C (a reference in the Acts of the council of 787 to the fact that a number of monasteries still remained in state hands); Nicephori patriarchae Antirrheticus iii adversus Constantinum Copronymum (in: PG 100, 376-533) 493D; Nicephori Patriarchae refutatio et eversio (in: P.J. Alexander, The Patriarch Nicephorus of Constantinople: Ecclesiastical Policy and Image Worship in the Byzantine Empire [Oxford 1958] 242-262[with comm. and lit. 180ff].), see 247.

[35] See R.M. Harrison, ed., Excavations at Saraçhane in Istanbul, I (Princeton 1986) 279-280 (M.F. Handy).

[36] See H. Ahrweiler, ‘Recherches sur l’administration de l’empire byzantin aux IXe-XIe siècles’, BCH 84 (1960) 1-109, at 55f.; R. Guilland, Recherches sur les institutions byzantines, 2 vols. (= BBA 35, Berlin-Amsterdam 1967) I, 428-430.

[37] Corpus Inscriptionum Graecarum, ed. A. Böckh (vols.II and II, I. Franz (vols. IIIff.) (Berlin 1828ff.), IV, xl, 315, no. 8680.

[38] Guilland, Recherches sur les institutions byzantines, I, 564f.; V. Laurent, Le Corpus des sceaux de l’empire byzantin, II: l’administration centrale (Paris 1981) 465ff.

[39] E.g. Haldon, ed., Constantine Porphyrogenitus, Three Treatises, (C) 742seq., 825seq. For the Hebdomon, see Janin, Constantinople byzantine, 446-449; Guilland, Études de topographie de Constantinople byzantine, II, 62.

[40] See Haldon, Byzantine Praetorians, 274-275; A. Pertusi, ‘Il preteso thema Bizantino di “Talaja”‘, BZ 49 (1956) 85-95.

[41] See the remarks of A. Toynbee, Constantine Porphyrogenitus and his World (London 1973) 220f.

[42] For the location of Malagina, on the Sangarios, see C. Foss, ‘Byzantine Malagina and the lower Sangarios’, Anatolian Studies 40, pp. 161-83, 1990 (repr. in Cities, fortresses and villages of Byzantine Asia Minor [Aldershot, 1996] no. VII). Following S. Şahin (Katalog der antiken Inschriften des Museums von Iznik [Nikaia] II, 3 [Bonn, 1987], p. 22f., and 150). Foss suggested that the base at Malagina was already established by the time of compilation of the so-called Apocalypse of Pseudo-Methodius, since it is mentioned there among the places at which an Arab force would winter in preparation for the attacks on Constantinople in the period 674-8 (A. Lolos, ed., Die Apokalypse des Ps.-Methodios [Meisenheim a. Glan 1976], xiii, 7). But this section of the text has been shown to be a much later interpolation (late eighth or early ninth century), so that this early date must be abandoned: see W.J. Aerts, ‘Zu einer neuen Ausgabe der ‘Revelationes’ des Pseudo-Methodius (syrisch-griechisch-lateinisch)’, in W. Diem & A. Falaturi (eds), XXIV. Deutscher Orientalistentag: ausgewählte Vorträge(Stuttgart, 1990), pp. 123-30, esp. 129-30.

[43] For convenience, see on Malagina The Oxford Dictionary of Byzantium, 3 vols., ed. A. Kazhdan et al. (Oxford 1991), 2, 1274. On the aplêkta see now Haldon, ed., Constantine Porphyrogenitus, Three Treatises, 62-64; and comm. to text (A) at 155ff. See also Toynbee, Constantine Porphyrogenitus and his Reign, 300ff.

[44] See F. Hild, Das byzantinische Straßensystem in Kappadokien (Vienna 1977) (= Veröffentlichungen der Kommission für die TIB 2 [Denkschriften der österr. Akad. d. Wiss., phil.-hist. Kl. 131]) 53; V. Aschoff, ‘Über den byzantinischen Feuertelegraphen und Leo den Mathematiker’, Abhandlungen und Berichte des Deutschen Museums 48 (1980), Nr. 1, 1-28; P. Pattenden, ‘The Byzantine Early Warning System’, Byz 53 (1983) 258-299; Haldon, ed., Constantine Porphyrogenitus, Three Treatises, 254-256.

This entry was posted in Articles and tagged , , , , . Bookmark the permalink.

Comments are closed.