Geoffrey Greatrex

title.none: Kaegi, Heraclius (Geoffrey Greatrex)

identifier.other: baj9928.0401.028 04.01.28

identifier.issn: 1096-746X

description.statementofresponsibility: Geoffrey Greatrex, University of Ottawa,

publisher.none: .

date.issued: 2004

identifier.citation: Kaegi, Walter E. Heraclius, Emperor of Byzantium. Cambridge: Cambridge University Press, 2003. Pp. xii, 359. $70.00 0-521-81459-6. ISBN: .

type.none: Review

relation.ispartof: The Medieval Review

The Medieval Review 04.01.28

Kaegi, Walter E. Heraclius, Emperor of Byzantium. Cambridge: Cambridge University Press, 2003. Pp. xii, 359. $70.00 0-521-81459-6. ISBN: .

Reviewed by:

Geoffrey Greatrex
University of Ottawa

The publication of a biography of the Emperor Heraclius is undoubtedly timely. Not only has there been no such work devoted to Heraclius since the start of the twentieth century, but, as Kaegi notes (10-16), there has been much discussion of his importance and achievements. Furthermore, new evidence and new ways of analysing the extant sources may now be brought to bear. As those familiar with Kaegi's work have come to expect, this biography takes into account, for the most part, the latest scholarship on Heraclius, including several works yet to be published; to his credit, Kaegi also makes good use of several relevant doctoral theses. Yet surprising lacunae remain, the most notable being the volume edited by G.J. Reinkink and B.H. Stolte, The Reign of Heraclius (610-641): Crisis and Confrontation (Louvain, 2002), the proceedings of a conference held in April 2001. Kaegi equally passes over the relevant volume in the series Histoire du Christianisme des origines a nos jours: volume 3, edited by L. Pietri, entitled Les églises d'orient et d'occident (Paris, 1998), covers the period with which the book is concerned. The reviewer may also be permitted to signal the absence of any reference to his own work (with S.N.C. Lieu), The Roman Eastern Frontier and the Persian Wars, A.D. 363-630 (henceforth REF) (London, 2002), which contains a large amount of material relevant to Heraclius' war against Persia.[[1]]

This review will go on to discuss in detail aspects Kaegi's work, but initially a few general points may be made. First, there is the question of the aim of the book. It belongs, clearly, to the genre of biography, a genre which continues to prove very popular. Within the late antique period, recent works such as N. Lenski, Failure of Empire: Valens and the Roman State in the Fourth Century A.D (Berkeley, 2002), S. Williams and G. Friell, Theodosius: The Empire at Bay (London, 1994) and J.A.S. Evans (The Age of Justinian: The Circumstances of Imperial Power (London, 1996) testify to this popularity.[[2]] But the titles or subtitles of these works are significant: all seek to offer a more rounded account of the period in question, concerning themselves not merely with the emperor but with the whole empire at the time. The emperor may loom too large in this picture sometimes -- such at any rate is the criticism of R. van Dam of Lenski's valuable work[[3]] -- but the approach is broad. Kaegi's work, by contrast, is merely subtitled "Emperor of Byzantium," and this accurately reflects his smaller focus. Given the very limited number of sources available for this period, a fact of which the author is all too conscious (7-14), this narrowness is puzzling. The reader searches in vain for overviews, for instance of the Roman empire at the opening of the seventh century, of the history of Romano-Persian relations, or of the doctrinal disputes to which Heraclius so unsuccessfully sought to put an end in the 630s.[[4]] To give one example, the first reference to Monothelitism in the work comes at p.209: "Sophronios had sworn an oath never to attack Monothelitism, and in fact he never explicitly attacked it." Neither here nor elsewhere is any attempt made to explain exactly what this doctrine was and why the newly appointed patriarch of Jerusalem was so fervently opposed to it. The work is consequently difficult to recommend for the lay reader with little or no previous background in Byzantine history. One should note, however, that some background information, at any rate on early Islam and the Byzantine state in the period of Heraclius, is to be found in Kaegi's preceding work, Byzantium and the Early Islamic Conquests (Cambridge, 1992). There are surprisingly few references to this useful earlier work in Heraclius, which is a shame, since it complements it well and helps to explain why certain subjects are passed over so briefly in the work under review.

It is perhaps a consequence of this more restricted focus on Heraclius that speculation abounds. A quotation suffices to give the tone: "There is no evidence that Heraclius ever wrote down any memorandum of advice for his immediate or future successors. Oral advice in the corridors of palaces is even more elusive to uncover. It is quite possible that he conferred extensively with Gregoria, his daughter-in-law, and left her detailed advice. It is unclear what insights and advice he passed on to any of them concerning imperial defenses in the east, the critical situation in Egypt, the Balkans, or theology. Presumably he urged them to follow the leadership of Philagrios in troubled fiscal matters and Patriarch Pyrrhos with respect to ecclesiastical issues. Unresolved, however, was how to reconcile competing fiscal needs to tap ecclesiastical wealth and ecclesiastical leaders' understandable resolve to preserve it untouched by bureaucratic predators. With respect to military crises, it is likewise uncertain how he counseled his successors to handle sensitive issues of military personalities and their rivalries and ambitions, as well as the pressures coming from units and men. It is unclear whether he advised these successors about the wisdom or urgency of campaigning in person. Likewise obscure is his advice on how to handle public opinion, whether in Constantinople or in the broader stretches of the empire." It becomes difficult sometimes to see the point of so much speculation, which, one may note, is particularly rife in the first chapter because so little is known of Heraclius' early life.

The book is marred by an excess of repetition: the same events are sometimes related (in almost identical terms) several times. A few examples may suffice: "Heraclius does appear to have been trying to stabilize the military situation southeast of the Tauros Mountains before proceeding to the Anatolian plateau and on to the Asian shores of the Bosphorus; he was not fleeing pell-mell after the defeat of his armies at the Yarmuk" (247), "In late 636, after the decisive destruction of much of his best armies at the hands of the Muslims at the battle of the Yarmuk, Heraclius did not take the shortest route out of Syria. Instead he chose to cover Armenia and his communications with Persia and with other friendly Arab tribes. His was no pell-mell retreat" (248).[[5]] A more serious example is the case of the Persian general Shahrvaraz, who briefly assumed the Persian throne in 630 before being assassinated. His seizure of the throne and Heraclius' support for him are initially discussed at p.188. No precise dates are offered in this context, however, nor at p.202, where Shahrvaraz's usurpation is again mentioned: the reader must wait until p.212 for these details, although here his reign is only mentioned in passing. The situation at the Persian court was certainly "fluid," as Kaegi puts it (202), but greater clarity could be achieved by a more unified treatment. One more important example should be noted. Once Heraclius had brought about the downfall of Khusro II in early 628, Khuro's son Kavadh II Shiroe succeeded to the throne and set about coming to terms with the Romans. Heraclius' subordinates proceeded to reclaim the cities of the Roman East which had been under Persian occupation. At Edessa, the emperor's brother Theodore had particular difficulty in persuading the Persian garrison to depart. The episode is recounted in this context by Kaegi at p.180, where he cites a relevant extract from Agapius. But then at p.203, Kaegi narrates the same events again, in much the same words. Indeed, the very same extract is cited, yet in a different translation. In this context, however, he argues that the events should be placed perhaps "as late as summer of 630, not the hitherto hypothesized 628 or 629." Rather in the style of the chronographers upon whom we depend so heavily, Kaegi has here created a doublet: he narrates the events in their usual place, implicitly dating them to 628, but then recounts them again in a new context. No argument is offered for the revision in the dating, which is in any case highly improbable, given that the sources in question refer specifically to Kavadh II, who died in October 628. There are also infelicities of style, sometimes to the point of obscuring the sense of a sentence. Such lapses should certainly have been picked up by an editor, and one would have expected a publisher of the stature of Cambridge University Press to have caught these slips.[[6]]

Despite all of these criticisms, some of which are relatively minor, the book certainly has its strongpoints. Kaegi's expertise in the analysis of military campaigns shines through on numerous occasions, well supported by good maps, plans and photographs (notably concerning the battle of Nineveh in December 627, pp. 158-68). His explanation of the manoeuvrings and counter-manoeuvres of Heraclius and the Persian army sent to pursue him under Razadh (Roch Vehan in the Armenian sources) is convincing; it is clear that Kaegi has benefited from his own observations in the region. Throughout his discussion of Heraclius' military operations, Kaegi well brings out the use of tactics described in the Strategikon ascribed to the Emperor Maurice. The bulk of the book is occupied with Heraclius' military campaigns, and so this strength is an important one, particularly since modern narrative accounts of the Persian war are few. One may note in passing, however, that in some cases disputes about points of historical, geographical or chronological detail are lightly passed over; some of these will be noted below, but one important instance must be cited here. Kaegi, while discussing Shahrvaraz's presence at Chalcedon during the siege of Constantinople in 626, accepts, after initial hesitation, that the general defected from the Persian cause to the Roman (148-50). Henceforth this supposition becomes the basis for further hypotheses: Heraclius, Kaegi suggests, might have benefited from information supplied by the renegade commander (152). He may even have decided to invade Persia from the north, Kaegi argues, in order not to pose a threat to Shahrvaraz (153). Further on, discussing the aftermath of the battle of Nineveh, Kaegi states that "Heraclius then encouraged the troops against Khusrau 'in order to frighten him', and summoned Shahrbaraz from Byzantium" (169). Shahrvaraz has thus now become subordinate to Heraclius. In fact, Kaegi's sentence is almost a precise translation of Theophanes' Chronographia here, save for the crucial last part: Mango and Scott's translation reads instead "After encouraging his army, the emperor pushed on against Chosroes with a view to frightening him and making him recall Sarbaros from Byzantium".[[7]] Despite Kaegi's presentation of events, scholars are far from united in accepting the notion of an alliance between Heraclius and Shahrvaraz, and it is a pity that he nowhere engages with those who dispute the historicity of the entente.[[8]] Indeed, it is striking that Kaegi rarely enters into discussions of points of detail or of controversies concerning particular topics; for the most part, he is content to offer a straight narrative, backed up here and there with references to recent work on the subject. However, particularly towards the end of the book, he does seek to argue against opposing views (e.g., on how much can be gleaned from Arab traditions about the early Islamic invasions of the Roman East, 239 n.22, 253 n.51). These all too brief discussions are stimulating and useful, and the reader can only regret that they are not more numerous.

It remains to discuss some of the points raised in the work in detail. The first chapter concerns Heraclius' Armenian background (itself problematic) and the period he spent in Africa during the exarchate of his father Heraclius. As already noted, because of the paucity of material, Kaegi is forced to indulge in considerable speculation in this chapter. There is useful information on Byzantine North Africa in the early seventh century; to the bibliography on this subject one might now add the recent volume of Antiquité Tardive 10 (2002), as well as Dossiers d'archéologie 268 (2001). Kaegi makes much of the importance of Heraclius' period spent in the province: "Heraclius' African years gave him a perspective that no other Byzantine emperor had, with the possible exception of some members of his own dynasty, because of his, their ancestor's, involvement in Africa" (34). This seems somewhat overstated. The Emperor Marcian, for instance, spent time in North Africa, some of it as a captive of the Vandals. It is even possible that Theodosius I -- the last emperor before Heraclius to take the field -- had also served in Africa; his father at any rate served there in the early 370s. There is no evidence in either case that a connection with Africa had any particular impact on the reign of the emperor in question. As Kaegi reaches the usurpation of Heraclius and his father towards the end of the first decade of the seventh century, more details emerge. The discussion of the mechanics of Heraclius' seizure of power is lucid and helpful, well drawing out the problems of legitimacy which would dog Heraclius' regime.[[9]]

Chapter two, entitled "Internal and external challenges in the first decade of the reign", covers the period from 610 to 620 in which the Persians took control of almost the entire Roman Near East. Kaegi rightly underlines the financial problems besetting the emperor, which became ever more acute as imperial territories diminished; he was therefore forced to raise funds from the church, a practice which both he and his subordinates would have recourse to on more than one occasion (60, 80-1, 86, cf. 272-5). The narrative of the Persian advance is satisfactory, although the effort required of the Persians to take control of the vastly enlarged Roman districts of Armenia (since the peace of 591) is not brought out (67). The reference to Theodosius, the supposed son of Maurice who had fled to the Persian court, at Edessa is also puzzling: although the former magister militum per Orientem Narses rebelled against the Emperor Phocas and held out against him in this city, there is no evidence whatever to suppose that Theodosius was proclaimed emperor there at this (or any other) point (67).[[10]] The narrative moves swiftly on to Heraclius' unsuccessful counter-attacks in the early 610s, perhaps too swiftly. The two invasions of Persarmenia by Philippicus are dealt with so cursorily as to leave it uncertain whether Kaegi believes that they should be identified. At p.75 he describes how Heraclius brought the aged general out of retirement and sent him on a mission eastwards, penetrating deep into parts of Armenia now taken over by the Persians; no date is offered for the episode, but the context suggests 613. Then at p.83 the reader is told, without further explanation, that "Philippikos, by making a diversionary raid into Persia, contributed to Shahin's lifting of the siege in unsuccessful pursuit of Philippikos." Shahin's attack on Chalcedon took place in 615. We are therefore dealing most probably with a second incursion by Philippicus, although this is far from apparent in Kaegi's account.[[11]]

Chapter three, "Taking the offensive", deals with the period leading up to Heraclius' counter-attacks in the 620s. Kaegi paints a vivid picture of the atmosphere in Constantinople at the opening of this critical decade, noting the ominous portents (such as an outbreak of plague in 619) which troubled its inhabitants and those of other parts of the empire (102-4). The transfer of the relics of holy men such as Theodore of Sykeon provided the emperor with a chance to restore morale and to boost support for his regime (105-6). Kaegi then narrates the controversial marriage of Heraclius to his niece Martina, a decision which he ascribes to a desire on the part of the emperor to assure the position of his dynasty (106-7).[[12]] The chapter closes with a description of Heraclius' initial campaign in 622. Kaegi places the emperor's victory over the Persians in Pontus, east of Euchaita, citing a passage from the Life of Theodore the Recruit; he also takes at face value the Expeditio Persica of George of Pisidia (115-116). In light of the arguments of James Howard-Johnston, such confidence is surprising; at the very least, some justification for pinning such faith in George of Pisidia is required, especially as the extract from the Life of Theodore the Recruit has also been associated with a Roman offensive in 626 rather than the events of 622.[[13]] Kaegi's chronology here is in any case confused. On p.115 he places the battle "late in autumn of 622," yet on the following page he states that "A threat from the west, probably from the Avars' growing menace in the Balkans, compelled Heraclius to return to Constantinople late that summer (622)."

The fourth chapter, "Peril and hope," continues the narrative of Heraclius' campaigns before his final invasion of Mesopotamia. Kaegi begins well by stressing the propaganda broadcast by Heraclius in advance of his operations, propaganda which is reflected in late sources, such as (Pseudo-)Sebeos (122-4). Following closely the detailed account of Theophanes, Kaegi relates the campaign launched by Heraclius in April 624. Here (127, cf. 130) he justifies his dependence on Theophanes' account by noting that the chronicler probably "drew this account from some offical or semi-official report issued by Heraclius"; this is certainly acceptable, although the absence of any reference to the important article of J. Howard-Johnston about this very point ("The Official History of Heraclius' Persian Campaigns" in E. Dabrowa, ed., The Roman and Byzantine Army in the East [Cracow, 1994], 57-87), alluded to elsewhere by Kaegi, is surprising here. In the section which follows, Kaegi follows the new chronology of Heraclius' movements put forward by Constantin Zuckerman ("Heraclius in 625," Revue des études byzantines 60 [2002], 189-97), which compresses events normally placed in 624 and 625 into the one year, 624 (128-30). Zuckerman has some good arguments in his favour, most notably certain inconsistencies in Theophanes' chronology, but the innocent reader of Kaegi's work would scarcely appreciate that there was any controversy whatever about the dating of these campaigns, still less that the orthodox view would place the outmanoeuvring by Heraclius of the three Persian general sent to defeat him in 625, rather than 624. Whatever the merits of Zuckerman's arguments, the point needs to be argued and justified; moreover, insufficient attention has been given to the difficulty of campaigning in winter in this mountainous and inhospitable region. The knock-on effects of Kaegi's adoption of Zuckerman's chronology are significant. Heraclius' withdrawal westwards and the battle at the river Sarus are accordingly placed in 625, rather than 626 (132). Kaegi then offers an account of the siege of Constantinople in 626, after which, in 627, the emperor proceeded to Lazica (142). An obvious problem presents itself therefore: what was Heraclius doing in 626? As is well known, the emperor was absent from Constantinople during the siege. We must suppose that he had good reason to stay away at such a critical moment. Yet on Kaegi's chronology, the emperor seems to have been almost entirely unoccupied during 626, save for some indirect contacts with the Kök Turks in the Caucasus. Something is awry here, and some sort of explanation is required.[[14]] Kaegi's discussion of the diplomatic successes of Heraclius in the Caucasus, notably in attracting the support of the Kök Turks, is valuable, highlighting the diplomatic skills exercised by the emperor (145-6). The last point we must note concerns the whereabouts of Shahrvaraz -- whose loyalties we have discussed above -- in the wake of the siege of Constantinople. Kaegi places him in Syria, "probably in Cilicia and in and around Alexandretta (Alexandria ad Issum) and Antioch." Most scholars, however, have supposed him to have been in Egypt, at Alexandria; Sebeos, at any rate, places him there at the fall of Khusro II (tr. Thomson and Howard-Johnston, The Armenian History attributed to Sebeos [Liverpool, 1999], 88). This is accepted by Howard-Johnston, "Heraclius' Persian Campaign," 28 and by Mango, "Deux études," 109. Again, it is incumbent on Kaegi to argue for a different location rather than simply to pass over earlier views.[[15]]

Chapter five, "The invasion of Mesopotamia", narrates the final, decisive campaign undertaken by Heraclius. Here Kaegi's account is at its best, particularly in relating the campaign of 627 and the manoeuvres to the north of Ctesiphon (as noted above). The maps and plans greatly enhance his account.[[16]] Problems arise, however, in dealing with the evens of December 627/January 628 and the flight of Khusro II. In late December, Heraclius was advancing ever closer to Dastagerd, Khusro's palace, and to Ctesiphon. Then, according to Kaegi, Khusro fled Dastagerd on 22 January; he cites B. Flusin, Saint Anastase le Perse et l'histoire de la Palestine au début du VII siècle (Paris, 1992), vol.2, 281 for this date. But as Howard-Johnston ("Heraclius' Persian Campaigns," 5 n.15) points out, this chronology is highly unsatisfactory, even if it does have the support of the Acts of Saint Anastasius: it becomes impossible to understand what Heraclius' army is doing during the month which intervenes, having sped southwards so swiftly in December. In fact, Kaegi is at least partly aware of these problems, for later on p.171 he reports that Armenian deserters informed Heraclius, soon after 1 January 628, that Khusro had already fled from Dastagerd. Thus Kaegi, having initially apparently favoured Flusin's chronology, has reverted to the traditional view. This is confirmed on the following page, where he places (following Theophanes' chronology) Heraclius at Dastagerd on 6 January.[[17]]

The narrative moves swiftly onwards, pausing only briefly to note the restoration of the frontiers of 591 by Khusro's successor Kavadh II Shiroe (179-80), closely followed by the king's death in October 628 (181). Persian affairs are dealt with somewhat piecemeal over the following pages, which makes them hard to comprehend: we hear of Ardashir's reign, which followed Kavadh II's, at pp. 184-5, and then of Shahrvaraz's seizure of power in April 630 and his three-month reign. There is further discussion of Shahrvaraz on pp. 188-9 and of Heraclius' plan to associate his family with that of the general by marriage, but the serious chronological difficulties which attend these events are passed over. In particular, no effort is made to explain what exactly Shahrvaraz's position was before ascending the throne in Persia; for it is known that he had returned there, with Heraclius' support, in late 629 or early 630. Some scholars believe that it was acting as a sort of regent for Ardashir that he returned the Cross to Heraclius. Since the Cross was restored to Jerusalem in March 630, i.e. before Shahrvaraz took over the throne, it is of some importance to understand what exactly the general's role was in the hand-over of the relic to the Romans and where precisely he wielded power. Kaegi seems to regard him as exercising control in Palestine and Syria until relatively late -- by implication until October or November 629 even (189). How he was simultaneously able to take power in Lower Mesopotamia is unclear and requires discussion.[[18]] One further important issue arises in this chapter: did the Persians ever capture the Holy Sponge and the Holy Lance? According to Kaegi (189), both these important relics were returned to the Romans and sent to Constantinople in late 629 by Shahrvaraz's son Nicetas. Kaegi takes a revisionist line, furthermore, with the despatch of the Holy Sponge and the Holy Lance to Constantinople (189): he follows a recent article by Holger Klein in supposing the Chronicon Paschale to have misplaced events connected with the arrival of these important relics in Constantinople. Instead of the conventional view, which has the general Niketas sending these items to Constantinople around the time of the Sasanian capture of Jerusalem (i.e. in 614), Klein argues that the Niketas concerned is the son of Shahrvaraz and that the relics actually reached the city rather in 629. Klein's case certainly deserves to be taken seriously, and it is to Kaegi's credit that he has made use of it; but it would have been a still greater service to the reader to be made aware of the debate here and of the fact that not every scholar has taken this line.[[19]]

Chapter six is entitled "Five crucial years: a narrow window of opportunity," and covers the period following Heraclius' stunning victory. Kaegi stresses the upheavals caused to the Roman empire by the years of warfare and Persian occupation, as well as the confusion occasioned by the displacement of populations (195, 207-8). But he also brings out the attempts made at rebuilding and restoration by the imperial government, putting to good use recent work, e.g. in Cyprus (209), where imperial building activity is attested. Much time is spent on working out the emperor's movements in the period (202, 209-10, 214); by contrast, little effort is made to explain the theological disputes which defied Heraclius' attempts to resolve them. The chapter also discusses the prevailing anti-Jewish sentiment of the period (205) and the beginnings of the Arab raids on southern Palestine (218). Kaegi considers briefly changes in imperial administration, notably the rise in importance of the sacellarius (227), which he connects with the financial problems of the empire (cf. 275). On the whole, however, Kaegi keeps discussion of administrative matters to a minimum, as he states clearly in his introduction: "Today there is much less interest in institutional history and institutional interpretations of the causes, nature, and markers of historical change and historical significance, irrespective of what role, if any, Heraclius had with any or all institutions" (11). Certainly, as Kaegi shows, there is no dearth of work dealing with administrative change in the seventh century.

The seventh chapter, "Tested again," deals with the Arab invasions of the Roman East. Kaegi approves of Heraclius' decision to remain in the East to confront this new challenge, seeing in it an attempt to understand the new adversary and to find ways to overcome him (234-7). The emperor refused to engage them in person, however, and his subordinates proved lacking. Kaegi likewise believes that the hard line taken by the emperor, refusing to countenance any negotiations with the Arabs, was justified (247-8, cf. 312), serving to stiffen resistance among the Romans, even if in the short term it led to the fall of many important provinces.

The eighth chapter, "Losing control," documents the closing years of Heraclius' reign and the dynastic problems which confronted the ageing emperor, although it is worth noting that the work concludes with Heraclius' death and offers no account of the troubled years which followed. Kaegi paints a gloomy picture of the mood in Constantinople as the empire in the East was steadily eroded; the emperor himself preferred to remain at Hiereia, on the other side of the Bosphorus (287-8). He had few advisers left on whom to depend (292), although Philagrius was clearly proving to be an effective sacellarius (272-6). One error may be noted here: while describing the rituals which took place in Constantinople during the period, Kaegi refers to the presence of Yazdin, Khusro II's important tax official (267). Here he has been led astray by the De Ceremoniis, which in fact just refers to patrikios ho kata Iesden. Yazdin himself was executed by Khusro before the end of his reign and thus cannot have taken part in any ceremony in Constantinople in the late 630s.[[20]]

The final chapter offers "Conclusions." Kaegi comes to the defence of Heraclius here, asserting that the empire was at any rate not much worse off than it had been when he assumed power in 610 (303). He suggests that too much money may have been spent on building and restoration projects in the wake of the triumph over the Persians (311), leaving insufficient funds to raise troops to deal with the new threat. Whatever his shortcomings, Heraclius' posthumous reputation remained considerable, as the anecdote about Constans II digging up his crown attests (320-1).

In sum, Kaegi's book has many merits. There are few modern accounts of Heraclius' Persian campaigns and none equipped with the useful maps and plans to be found here. His knowledge of the primary and secondary sources is good, even if there are some important lacunae, signalled above. For the military historian, this is certainly a useful volume. For the general student of Roman or Byzantine history, it will need to be supplemented by other accounts, dealing (for instance) with institutional history (consciously left to one side by Kaegi) and with church affairs. It is to be hoped that a revised edition will be brought out by Cambridge University Press in paperback.


[[1]] A few other minor omissions from the bibliography deserve mention. Since Kaegi spends some time discussing St Theodore of Sykeon (e.g. 9-10, 75-6), one would expect to find some reference to the extensive treatment of the saint and his milieu in S. Mitchell, Anatolia. Land, Men and Gods in Asia Minor (Oxford, 1993), vol.2, 122-50. An article of Benjamin Isaac, "The Army in the Late Roman East: the Persian Wars and the Defence of the Byzantine Provinces" in A. Cameron, ed., The Byzantine and Early Islamic Near East. States, Resources, Armies, Studies in Late Antiquity and Early Islam, vol.3 (Princeton, 1995), 125-55 is also unfortunately absent from Kaegi's work: a substantial part of this article argues that Heraclius deliberately neglected the eastern provinces and concentrated instead on devastating the Persian heartlands (esp. 133-7). His argument is unconvincing, given that the swift victory of Heraclius by his final campaigns soon restored the provinces to Roman control, but it deserves consideration nonetheless.

[[2]] See my review of Friell and Williams' book in Journal of Ecclesiastical History 48 (1997), 136-7 and of Evans' book in Classical Review 48 (1998), 404-6.

[[3]] R. van Dam, review of Lenski in BMCR 2003.07.44.

[[4]] One will find such an overview in the stimulating work of M. Whittow, The Making of Orthodox Byzantium, 600-1025 (London, 1996), as also in the useful article of J. Howard-Johnston, "Heraclius Persian Campaigns and the Revival of the East Roman Empire, 622-630", War in History6 (1999), 1-44. There are brief glances at events taking place in the empire during Heraclius' reign, e.g. in Spain (89, 141), and yet odd gaps remain. Little space is devoted to the Persian occupation of Egypt (91-2), for instance, or to the feints by Philippicus into Persian territory (75).

[[5]] Cf. pp.138 and 141 on possible contacts between Avars and Persians.

[[6]] A few examples may be cited. "Khusrau II raised troops to halt the Byzantines, after great loss of life, because of their hasty pursuit and lack of provisions, they subjugated the territory that Philippikos had traversed" (75). It is unclear just who is pursuing whom. Referring to Egypt, Kaegi offers the following analysis (211): "Although an extremely valuable province with respect to agriculture and tax revenues, and although it was wracked with religious dissensions and had suffered at the hands of its recent Persian occupiers, Heraclius did not turn south from Jerusalem, via land or sea, to visit Egypt." The lack of editorial attention emerges equally from minor errors in footnotes: at 290 n.70 the footnote ends with a completely irrelevant sentence concerning the tomb of Heraclius. It actually belongs at 320 n.24, where it may also be found. Note also "Atotrapene" for "Atropatene" at p. 177; the same sentence appears in n.50 on p.38 and in the main text. There is some confusion at p. 41 over the coin depicting Heraclius and his father: the caption beneath the solidus states (correctly) that the Emperor Heraclius is on the left, while his father is on the right. In the text below, however, Kaegi says that the coin represents the Emperor Heraclius on his father's left; clearly, this is the opposite of what is said in the caption.

[[7]] C. Mango and R. Scott, The Chronicle of Theophanes Confessor (Oxford, 1997), 450, cf. Greatrex and Lieu, REF, 214. Interestingly, Mango and Scott note (455 n.8) that Anastasius' translation of Theophanes omits the reference to Shahrvaraz entirely. The quotation from Sebeos at p. 153, according to which Heraclius "was troubled by fear of Khoream [Shahrbaraz]" also casts doubt on Kaegi's interpretation: if there was an agreement between the two men, such fear would be misplaced.

[[8]] See Howard-Johnston, "Heraclius Persian Campaigns," 22 n.68, against the genuineness of the treaty with Greatrex and Lieu, REF, 205.

[[9]] We may note here a few more minor errors. Kaegi claims (25) that Solomon, the sixth-century general, was from Armenia; in fact, cf. The Prosopography of the Later Roman Empire 3, ed. J. Martindale (Cambridge, 1992), Solomon, he was from near Dara. At p. 39 Kaegi places the fall of Dara to the Persians in 605; at p. 67 he somewhat less confidently states that Dara fell "possibly in 605". In fact, 604 is much more likely: see Greatrex and Lieu, REF, 184. At p. 42 Kaegi uses the term consularius when presumably he is referring to a consularis.

[[10]] See Greatrex and Lieu, REF, 183-4.

[[11]] For details see Greatrex and Lieu, REF, 189, 194, and 304 n.89 (noting that some scholars prefer to identify the two campaigns).

[[12]] Here it might have been opportune to note the arguments of C. Mango, "Deux études sur Byzance et la Perse sassanide," Travaux et Memoires 9 (1985), 114, who sees this marriage (as well as that of Heraclius' son Constantine to his cousin Gregoria) as stemming from close-kin marriage practices common in Mesopotamia and Osrhoene: Mango believes that it is likely that the emperor came from this region, rather than Armenia or Cappadocia. Kaegi dismisses Mango's arguments briefly at 21 n.4 and henceforth assumes that Heraclius was from Armenia.

[[13]] See the arguments of Howard-Johnston, "Heraclius' Persian Campaigns," 3 n.11. Kaegi follows Oikonomides closely, failing to note that others (cf. Greatrex and Lieu, REF, 307 n.14) have doubted whether George's account can be interpreted literally. Howard-Johnston's argument -- that a propagandist such as George would surely have highlighted the fact that the emperor had penetrated so deep into enemy territory, if such had been the case -- is persuasive, and George's silence on this point must be explained if one is to follow Kaegi's line. For the alternative dating of the battle near Euchaita see Greatrex and Lieu, REF, 207 (following Howard-Johnston).

[[14]] For the conventional chronology, cf. Greatrex and Lieu, REF, 202-9.

[[15]] In fact, some arguments are advanced on p.181, although they are essentially limited to noting that the sources are ambiguous as to whether Alexandretta or Alexandria is meant.

[[16]] Although Kaegi's bibliography is usually complete, one notes here the absence of V. Minorsky, "Roman and Byzantine Campaigns in Atropatene," Bulletin of the School of Oriental and Arican Studies 11 (1943-6), 243-65.

[[17]] For further details see Greatrex and Lieu, REF, 215-16 with nn.92, 98 (on the date of Khusro's flight). Flusin, ch.9, esp. 268-70, examines the issue in detail, preferring the dating of the Acts to that of Theophanes. His case is a strong one, but in a work focussed on military history, it is all the more vital for Kaegi to examine the issue closely and to explain the sudden change of pace of Heraclius' movements, if Flusin's chronology is accepted.

[[18]] On these events see Greatrex and Lieu, REF, 227-8 (with bibliographical references to the controversy), cf. Howard-Johnston, "Heraclius' Persian Campaigns," 28.

[[19]] See, e.g., B. Flusin, Saint Anastase le Perse et l'histoire de la Palestine au début du VII siècle (Paris, 1992), vol.2, 180-1, for the conventional view; also Greatrex and Lieu, REF, 191-2. Cf. H. Klein, "Niketas und das wahre Kreuz. Kritische Anmerkungen zur Ueberlieferung des Chronicon Paschale ad annum 614," BZ 94/2 (2001), 580-7. In Klein's favour is the fact that the day and date offered by the Chronicon Paschale (ed. Dindorf, p. 705) under 614 do not correspond: in that year the 28th October was not a Saturday (the day the Holy Lance arrived in Constantinople). Such discrepancies are present elsewhere in the Chronicon, however. The other arguments advanced by Klein, connected with the development of ceremonies connected with the Cross in Constantinople, are more tenuous, especially as the difference between the two dates in question is only 15 years. One counter-argument may be noted here, in favour of retaining the Chronicon's existing chronology. In describing how the Romans obtained the Holy Lance, the author notes that it was handed over to the Romans by one of Shahrvaraz's entourage, "after it (i.e. Jerusalem) had been captured by them" (tr. M. and M. Whitby, Chronicon Paschale, 284-628 AD [Liverpool, 1989], 157); and this person then passed it on to Niketas. Neither in this entry, nor in the previous one (concerning the arrival of the Holy Sponge in Constantinople) is there an indication of who the "they" are: one must go back to the preceding entry, concerning the fall of Jerusalem, to realise that the "they" are, of course, the Persians. Thus the three entries would seem to fit together. This is not a conclusive argument, of course, but it is doubtful whether Klein's redating of this episode will become the orthodox view, any more than earlier such attempts to rearrange events towards the end of the Chronicon Paschale.

[[20]] See PLRE III, Anonymus 2 for this person, present at Hagia Sophia on 1 January 639. The reference to Yazdin on this page of Kaegi's work is omitted from the index.