Hendy Byz Econ 1081-1204 Reappraisal

Byzantium, 1081-1204: An Economic ReappraisalAuthor(s): M. F. HendyReviewed work(s):Source: Transactions of the Royal Historical Society, Fifth Series, Vol. 20 (1970), pp. 31-52Published by: Royal Historical SocietyStable URL: http://www.jstor.org/stable/3678761 .Accessed: 29/11/2011 09:36

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

Royal Historical Society is collaborating with JSTOR to digitize, preserve and extend access to Transactions ofthe Royal Historical Society.

http://www.jstor.org

BYZANTIUM, Io8I-I204: AN ECONOMIC REAPPRAISAL

By M. F. Hendy, M.A.

READ 14 FEBRUARY I969

HE Byzantine Empire of the period Io8I-I204 is gener- ally considered to have been culturally brilliant but

economically decadent. The standard against which this decadence is measured is the situation supposed to have existed

during the ninth and tenth centuries when the Empire consisted

basically of the Balkan coastlands, the Aegean islands and Asia

Minor, and when it possessed a flourishing agriculture dependent upon a free peasantry which also supplied the manpower of its

army and navy, a vital urban life, and control of its extensive internal and external trade. Its revenue was therefore assured and its coinage stable. By the twelfth century it had lost the greater part of Asia Minor which had formed the factor essential to its

agricultural, military and urban life. The first two were now

largely in the hands of feudal magnates who commanded ruin-

ously expensive but unreliable mercenaries, and the trade of the

Empire had fallen under the control of the Italian merchant cities. The reduced revenue was incapable of standing the strain placed upon it by increased expenses, the difference being made up by the debasement of the coinage-which caused further chaos in economic life.l

A picture that is at once composite and simplified but, it is

hoped, also recognizable. While some of its elements are undoubt-

edly correct, however, the inferences drawn from several seem

1 The economic history of the Byzantine Empire lacks adequate general treatment. The following works may be consulted: S. Runciman, Byzantine Civilization (London, I933), pp. I63-222; A. Andreades in Byzantium: an Introduction to East Roman Civilijation, eds. N. H. Baynes and H. St. L. B. Moss (Oxford, 1948), pp. 51-70, 7I-85; S. Runciman in The Cambridge Economic History of Europe, ii eds. M. Postan and E. E. Rich (Cambridge, I952), pp. 86-II8. Most general histories include some commentary upon economic affairs, and particularly relevant to this paper seems to be: G. Ostrogorsky, History of the Byzantine State, trans. J. Hussey (Oxford, 2nd edn, I968), pp. 357, 369-72, 374, 393-94.

31

32 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

of dubious validity and others are demonstrably incorrect in both respects.

The major difficulty confronting the economic historian of the Byzantine Empire is an almost complete lack both of official and even of private records, a problem which, admittedly, is shared by western mediaevalists but which for them is increas- ingly alleviated (if never to the extent they would wish) from the twelfth century onwards. The result is that the economic history of the Empire is throughout heavily based on what can be gleaned from the occasional reference in chronicles, hagiographies, legal codes, a few monastic archives and descriptions by foreign travellers. There is no possibility of quantification and the detec- tion of whole movements must rest on the chance evidence of individuals who not only demonstrated a perverse ability to misunderstand (not to say flagrantly misrepresent) but who were notoriously capable of considering information about the economy of their state a pain to record and an imposition upon their readers.

The first part of this paper will therefore be devoted to an attempt at a conspectus of the salient features of the economic situation during the twelfth century using the sources described above and including criticisms of fact and inference in accepted views where they seem necessary. The second will bring the evidence of archaeology and the imperial coinage (the two orders of evidence that permit comparative estimation) to bear upon the problem of the urban development of the Empire, and utilize recently published evidence to confirm what appears to be their consistent conclusion.

Since the ninth and tenth centuries the Empire had lost its southern Italian provinces and the central plateau of Asia Minor. It had, on the other hand, gained the inner Balkans (finally since IoI8 and until the late II8os), Crete (since 96I), and Cyprus (since 965 and until II9I). It had apparently retained some form of control over the Crimea and adjacent southern Russia,1 and it

1 The latest treatment seems conclusive: G. G. Litavrin, 'A propos de Tmutorokan', Byjantion, xxxv (i965), pp. 221-34.

BYZANTIUM, I08I-1204: AN ECONOMIC REAPPRAISAL 33

had both won and lost Antioch (969/I085). The overall terri- torial concentration had therefore tended to shift away from Asia in favour of Europe, but what this represented economically remains completely uncertain. For whereas one might have expected some discussion as to the economic significance of the absorption of the Balkans north of the Rhodope into the Empire, it has instead almost exclusively revolved round that of the loss of the central plateau of Asia Minor.

The area of Asia Minor remaining under direct Byzantine control during the major part of the twelfth century comprised the coastlands, from Trebizond in the north-east round to Attalia in the south-west, and the river-valleys of the west. The western frontier with the Seljuks, if frontier it may be termed, left within the Empire Malagina, Philadelphia, Laodicea and Chonae; but between these cities, which were more or less permanently imperial, and those usually in various Turkish hands lay a broad band of disputed territory, roughly marking the perimeter of the central plateau and including such cities as Castamenon, Gangra, Claudiopolis, Dorylaeum, Cotyaeum, Sozopolis and Sublaeum. Comnenian military activity therefore seems to have followed the strategy of securing the perimeter and fortifying the territory behind it.' This implied distinction may well have had economic significance.

It should be emphasized that the physical structure of Asia Minor, involving a basic division into coastal plain and river- valley over against central plateau, is one which has clear-cut economic repercussions. Comparison of a map of physical struc- ture with maps of density of population, annual precipitation, natural vegetation, land-use and agricultural production, will illustrate this point most satisfactorily.2 Those areas that are now

1 An attempt has been made to illustrate the approximate extent of Comnenian territory in Asia Minor in a map to appear in: M. F. Hendy, Coinage and Money in the Byrantine Empire zo8z-z226 (Dumbarton Oaks Studies, xii) in press. This is largely based on the chroniclers Nicetas Choni- ates and John Cinnamus, and P. Charanis, 'On the Asiatic Frontiers of the Empire of Nicaea', Orientalia Christiana Periodica, xiii (1947), pp. 58-62. For the construction of fortresses see: H. Ahrweiler-Glykatzi, 'Les forteresses construites en Asie Mineure face a l'invasion seldjoucide', Akten des XI internationalen Byrantinistenkongresses (Munich, 1960), pp. 182-89.

2 E.g. Atlas of the Arab World and the Middle East (London, I960),

pp. 6, 38-40.

34 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

the most densely populated, the most well-watered, with the most favourable natural vegetation and allowing the most profit- able and varied use of land, are very heavily concentrated towards the periphery of the peninsula-in other words towards the coast-lands and river-valleys, which are precisely those areas that were then held by the Comneni. The implication is obvious, and although it would perhaps not be possible in any extensive or absolute way to prove that the economic character of the penin- sula in the Byzantine period paralleled that of today, the proba- bility that it did is very great.

But if this parallel is accepted, at what stage did it come into being? For although much of it is dependent upon natural phenomena that are unlikely to have changed radically, it was at some stage at least less close. A decline in the quality and density of Anatolian agricultural and urban life since ancient times is demonstrable. The customary answer would, of course, be that it must have occurred subsequent to the Seljuk invasions of the Io70s.1 But one may be permitted to doubt this.

In I955 W. C. Brice (a historian as well as a geographer), in an article which seems to have remained unknown to Byzantinists,2 brought forward evidence with which he sought to prove that the techniques of Anatolian agriculture would have tended to decline with the graecization and romanization of archaic Phry- gian society, and that the deterioration of the landscape had com- menced shortly before the opening of the Christian era. He pointed out that the features that characterize this deterioration; the denudation of topsoil; recession of forests and gradual desic- cation, must be geologically connected with the equally character- istic silting up of river-mouths and conversion of fertile lowlands into malarial swamp that is to be found at several places along the coast. Historical evidence quite clearly demonstrates that these processes had been already well advanced by the early Byzantine period, the immediate cause having perhaps been the reckless spread of large-scale commercial farming during the Hellenistic and Roman periods.

Documentary evidence for the urban history of later Byzantine 1 E.g. J. Laurent, Byrance et les turcs seldjoucides dans l'Asie occidentale

jusqu'en zo8z (Nancy, I913), pp. I02-09. 2 W. C. Brice, 'The Turkish Colonization of Anatolia', Bulletin of the

John Rylands Library, xxxviii (1955-56), pp. 18-44.

BYZANTIUM, IO8I--I204: AN ECONOMIC REAPPRAISAL 35

Anatolia is sparse and curiously contradictory.l On the one hand the Notitiae Episcopatuum of the eighth and ninth centuries seem to exhibit continuity and even expansion in the number of sees. This is equated by Ostrogorsky with an increase in the number of cities,2 but even if such a direct relationship were acceptable on other grounds (and it was not due, for instance, to a general reduction in the size of sees and a consequent increase in their number) it is considerably damaged by the iconodule accusation that Constantine V created new sees in order to pack them with supporters of his own religious persuasion.3 And on the other hand, as Charanis has observed,4 all that is known of the general facts of the contemporary situation seems to require a drastic reduction of both rural and urban life: at least two devastating bouts of plague, under Justinian I and Constantine V; Persian invasions under Phocas and Heraclius; continuous and deeply penetrating Arab invasions from the mid-sixth to the mid-ninth century and, in addition, a destructive civil war (that of Thomas the Slav) under Michael II.

It is in this context that the evidence of two ninth-century Arabic sources should be read. The first states that: 'In the days of old cities were numerous in Rum [Anatolia] but now they have become few. Most of the districts are prosperous and pleasant

1 The state of Byzantine cities in general during the seventh to ninth centuries forms a problem that has already given rise to a number of articles. A list of the main ones should include: A. P. Kazhdan, 'Vizantiiskie goroda v VII-XI vekakh', Sovetskaya Arkheologiya, xxi (I954), pp. 164-83; P. Charanis, 'The Significance of Coins as Evidence for the History of Athens and Corinth in the Seventh and Eighth Centuries', Historia, iv (I955), pp. 163-72; G. Ostrogorsky, 'Byzantine Cities in the Early Middle Ages', Dumbarton Oaks Papers, xiii (959), pp. 47-66; the various papers and commentaries in Actes du XIIe Congres International d'ttudes By;antines, i (Belgrade, 1963), pp. 1-44, 275-98; S. Vryonis, 'An Attic Hoard of Byzan- tine Gold Coins (668-741) from the Thomas Whittemore Collection and the Numismatic Evidence for the Urban History of Byzantium', Recueil des travaux de l'Institut d'Studes byrantines, viii (Belgrade, I963) (Melanges G. Ostrogorsky, i), pp. 291-300; E. Frances, 'La ville byzantine et la monnaie aux VIIe-VIIIe siecles', By~antinobulgarica, ii (1966), pp. 3-14. The use made of numismatic material leaves much to be desired.

2 Ostrogorsky, 'Byzantine Cities', pp. 52-61. 3 Ibid., p. 59; Frances, 'La ville byzantine', p. 4. 4 P. Charanis, 'Observations on the Demography of the Byzantine Em-

pire', Proceedings of the XIIIth International Congress of Byzantine Studies (Oxford, 1967), Main Paper xiv, pp. 454--59.

36 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

and have each an extremely strong fortress, on account of the frequency of the raids which the fighters for the faith direct upon them. To each village appertains a castle where in times of flight they may take shelter.'l Ibn Kordadhbeh also appears to make a distinction between 'cities', of which he names only a few, and 'fortresses' or 'fortified places' of which there were a much larger number.2

It may well be that in the final analysis of mid-Byzantine Anatolia we must envisage a small number of mainly coastal cities (such as Trebizond, Smyrna and Attalia) and a larger one of fortresses or fortified townlets, each the focal point of a rela- tively restricted rural area. Such a system would tend to be geared to meet military, administrative and local agricultural needs and would afford little opportunity for, or means of, industrial and mercantile development. Indeed, the basic unit of middle Byzantine administration, the large, essentially military theme, is not one that is easily reconcilable with an advanced urban economy.

In summary, therefore, even allowing for a degree of urban recovery during the tenth and eleventh centuries, the supposedly crucial role of central Anatolia in the economic life of the Empire should be viewed with suspicion. Such evidence as there is points to it having been of value for its agricultural rather than its urban life, and even the former may have been conducted at a lower level than previously, the main emphasis in both respects having shifted to the coastlands and river-valleys that remained in the possession of the Comneni throughout the twelfth century.3

Certain aspects of the civil, military and administrative systems of the Comneni and Angeli, against the background of which any evaluation of the economy must take place, have been the subject of studies by Ahrweiler and Lemerle.

The army contained an appreciable proportion of mercenary 1 Hydad al 'Xlam: The Regions of the World, trans. V. Minorsky (Oxford,

1937), P. 157. 2 Ibn Khordadhbeh, trans. M. J. de Goeje (Leiden, 1889), pp. 77-80. 3 Certainly this tendency seems already present in a list of major Anatolian

cities occurring in Theophanes' chronicle drawn up by Ostrogorsky in his article on Byzantine cities (pp. 61-62, note 64). Of the total of 34 cities usually in Byzantine hands during the seventh to ninth centuries no less than 20 are definitely in the coastal plain, and only 9 definitely on the plateau, many of these having obvious military functions. Five form marginal cases.

BYZANTIUM, 1081-1204: AN ECONOMIC REAPPRAISAL 37

elements, a fact commented upon by both mediaeval1 and modern authors, although one may suspect that numbers of them were hired for particular campaigns and Nicetas Choniates records in a general (and disapproving) way that the Byzantines themselves increasingly came to find a military career attractive.2

The navy, now the better known of the two military services, seems to have reached its twelfth-century apogee under Manuel I (1143-80) and from then on to have been neglected.3

It is, however, neither the composition nor even the degree of success achieved by the military services that is of economic interest, although one may in passing note that there seems no general reason why mercenaries should in themselves be less efficient and reliable than native troops, and no particular indica- tion that they were so during the course of the twelfth century. In fact, the principal economic problem involves the overall cost of the military services, either absolutely or in comparison with that of previous centuries, and the methods employed in payment -the distinction lying between the regular cash salary and the grant of land accompanied by the right to taxes accruing from it (the pronoia grant). Concerning their absolute or even com- parative cost it would be best to admit that no evidence exists, although the accepted supposition is that since mercenaries must be more expensive than native troops the comparative cost had

1 For instance, Odo of Deuil, De Profectione Ludovici VII in Orientem, ed. V. G. Berry (Records of Civilization, Sources and Studies, xlii [New York, I948]), p. 88. Byzantine sources record Russians, Varangians, English, French, Germans, Bulgarians, Turks, Alans, Abasgi and others-Actes de Lavra, eds. G. Rouillard and P. Collomp, i (Paris, I937), no. 41, p. III; Acta et Diplomata Graeca Medii Aevi, eds. F. Miklosich and G. Miller, vi (Vienna, 1890), p. 47.

2 Nicetas Choniates, Bonn edn, p. 273. Concerning the size of the con- temporary Byzantine army virtually nothing is known: the Emperor Manuel, writing to Henry II of England, records that the army taken on the Myrio- cephalum campaign, and including the baggage and siege train, stretched out over ten miles when moving in file owing to the difficulties of the terrain- but this is incapable of verification. See A. A. Vasiliev, 'Manuel Comnenus and Henry Plantagenet', By{antinische Zeitschrift, xxix (1929/30), pp. 237-38.

3 H. Ahrweiler, By5ance et la mer (Paris, 1966), particularly pp. I75-297. An interesting description of the fleet sent against Alexandria by Manuel, in I I69, and recorded by William of Tyre, confirms the strength and elabora- tion of contemporary Byzantine naval forces: Historia Rerum in Partibus Transmarinis Gestarum, 20, xiii (Recueil des historiens des croisades: historiens occidentaux, t. i, pt. ii [Paris, 1844], p. 96I).

38 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

risen. Even the extent to which the pronoia grant was utilized at this comparatively early stage of its existence is a matter for dis- cussion: Ostrogorsky would seem to regard it as in widespread use already under Alexius I (Io8I-III8), but with a significant expansion under Manuel;' Lemerle would doubt its predominance even under the latter Emperor.2 The crucial question is, of course, whether the military services rendered by the grantees adequately compensated the state for the surrender of its rights of taxation: if they did, then, however socially unpopular, the institution forms merely another method of payment and is devoid of immediate economic consequence. Since the grant was as yet not hereditary, remained inalienable, and was strictly controlled in size,3 the balance may not have been so overtly unfavourable.

The revival of powerful armed forces by the Comneni was accompanied by a reform of the machinery of civil and military administration.4 To the extent that its end was the more complete integration of the economic and military resources of the Empire it contrasted with the policy of the previous regime, but the means employed to achieve this end, centralization, was very much a feature of the preceding period. And it was surely in this policy of centralization, rather than in any immediate economic event or trend, that lay the greatest danger to the Comnenian reconstruction of the Empire. For it was fundamentally incom- patible with the major social phenomenon of the times: the growth of the great territorial interest.5 Now it is by no means

1 G. Ostrogorsky, Pour l'histoire de lafeodalite byzantine (Brussels, 1954), pp. 26-54.

2 P. Lemerle, 'Recherches sur le regime agraire a Byzance: la terre militaire a l'epoque des Comnenes', Cahiers de civilisation medievale Xe-XIIIe siecles, ii (I959), pp. 265-81.

3 G. Ostrogorsky, Quelques problemes d'histoire de la paysannerie byrantine (Brussels, 1956), pp. 25-40.

4 Typified by the strengthening of the position of the logothetes ton sekreton, controlling and coordinating all the various governmental depart- ments; the creation of the post of megas logariastes, coordinating the genikon and stratiotikon-the financial and military departments; the unification of military and naval command under the megas domestikos and megas doux respectively; the eventual re-establishment of a themal system, each under the short-term civil and military control of its doux kai anagrapheus. See: Ahrweiler, Byjance et la mer, pp. 200-I I, 272-79. 5 There should be a close connexion between the growth of the territorial interest and the rapidly developing sense of family and descent as expressed

BYZANTIUM, Io8I-I204: AN ECONOMIC REAPPRAISAL 39

obvious that the territorial interest as evolved in Byzantium was an immediate economic disadvantage in itself-agricultural efficiency is one possible beneficiary, and surplus produce has to be sold in any case-but the potential political and economic threat to the central government is clear. Although there seems to be no particular evidence that the Comnenian emperors were actively threatened in either way, once the dynasty had estab- lished itself, the incompatibility would have remained and even increased. The days when the momentum of government could be kept up by the system alone without constant and consistent direction from above had long passed, and although the personal qualities of Alexius, John and Manuel were sufficient to en- sure its maintenance the inherent crisis very quickly followed the death of the last of them, in i I80. When the course of events in time-and by chance-threw up an emperor (Isaac II Angelus, I 185-95) who was of a family of comparatively minor and recent eminence, and not himselfa strong character, the conflict of interests became blatant and active. The machinery of government re- mained, but its proper functioning was now continually thwarted.1

The most important feature of the twelfth century, with regard to the Mediterranean as a whole, is customarily-and correctly- considered to have been the emergence of the Christian countries of its western half into a stage of development where they were in a position to play an active and even decisive political and economic role; hitherto the prerogative of the Christian and Muslim countries of its eastern half.

But the speed with which the eventual western economic stranglehold was established (its political initiatives were far less successful) should not be exaggerated. There is no evidence, as sometimes alleged,2 that a shift in trade-routes began to deprive the Empire of the traffic that directly and indirectly provided it with what must have been an important element of its total revenue. By its very geographical situation Constantinople

in the evolution of multiple surnames. The course of the latter is well illustrated in: D. I. Polemis, The Doukai, a Contribution to Byzantine Prosopography (London, I968).

1 For example, P. Lemerle, 'Notes sur l'administration byzantine a la veille de la IVe croisade d'apres deux documents inedits des archives de Lavra', Revue des etudes byzantines, xix (1961), pp. 258-72.

2 Runciman, Byzantine Civilization, p. I69.

40 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

straddled a main north-south trade axis, and as long as the Empire held most or all of the strategic ports of southern Asia Minor, Cyprus, Crete and the southern Peloponnese, it stood to derive similar benefits from the main east-west one. The testi- mony of the mid-twelfth-century Spanish Jew, Benjamin of Tudela, and the more or less contemporary satire Timarion confirm, if confirmation was needed, the continuing international status of Constantinople and Thessalonica.1

It may be (although no figures exist) that the relative propor- tion of Mediterranean trade that was carried by Byzantine ship- ping had declined since the ninth and tenth centuries, when the Empire had been further advanced economically than most of its neighbours, and all of its western and northern ones. But it is surely an adventurous proposition that because it was not now capable of both producing raw materials and manufactured goods, and providing the bottoms in which the bulk of them were exported, it therefore suffered economic harm. Despite the assumption that the Byzantine merchant class was in decline, twelfth-century documentary evidence indicates the presence of its members at both ends of the Mediterranean: Alexandria, Barcelona and Beziers.2 Evidence for Byzantine shipping tends to be confined, by the nature of its largely monastic origins, to indicating the existence of an internal marine.3

The concession of favourable customs-rates to western mer- chants (immunity for Venetians, 4 per cent for Pisans and Geno- ese, in place of the normal io per cent) has occasioned a good deal of unfavourable comment, on the grounds that it must have involved considerable losses to imperial revenue. The situation seems more complex. Concessions were still, at this period, fre- quently subject to restriction-either geographically, for instance

1 The Itinerary of Benjamin of Tudela, trans. M. N. Adler (London, 1907),

pp. 12, I3; 'Timario sive de Passionibus ejus', ed. M. Hase in Notices et extraits des manuscrits de la Bibliotheque Imperiale, et autres bibliotheques (Paris, 1813), pp. I7I-73 (2e partie).

2 Itinerary, trans. Adler, pp. 2, 3, 76; Documenti sulle relaZione delle cittd toscane coll' Oriente Cristiano e coi Turchi, ed. G. Miiller (Florence, I879), no. 41, pp. 66-67; S. D. Goitein, A Mediterranean Society, i, Economic Foundations (University of California, 1967), pp. 44-46.

3 There is, for example, a useful list of fiscal immunities accorded the ships of various monasteries in H. Antoniadis-Bibicou, Etudes d'histoire maritime de ByZance d propos du 'Theme des Caravisiens' (Paris, 1966), pp. I32-33.

BYZANTIUM, IO8I-I204: AN ECONOMIC REAPPRAISAL 41

excluding Cyprus or the Black Sea, or in distinguishing between internal trading and exportation.' Most were granted compara- tively early (1082 for Venice, I I I I for Pisa, 115 for Genoa), and since the overall volume of east-west trade increased very con- siderably in the course of the twelfth century2 it would seem logical to suppose that the long-term increase would have tended to compensate for the short-term loss, at least in the case of Pisa and Genoa. In the case of Venice it should be observed that the residence and operation of numbers of foreign merchants within the Empire are likely to have entailed the kind of benefit that is difficult to measure at the best of times: the investment of capital and reinvestment of profit. Since customs concessions themselves tend to encourage trade it should also be observed that any increase in demand for Byzantine products on the part of the west should have had the result of either stimulating Byzantine production or of increasing the prices that they were able to charge-or of course, both. As to the actual balance of trade, it is supposed by Lopez to have remained favourable to the Byzan- tines throughout the twelfth century.3

None of these trends, it must be added, is capable of proof or disproof by figures, but given the known circumstances some or all of them ought to have occurred, and they should at least be kept in mind when attempting to evaluate the effect of western involvement in Byzantine economic life, particularly before the establishment of physically separate and self-sufficient colonies such as those set up by Venice after I204 and by Genoa after I261.

II

By this stage it must have become quite clear that the sources of evidence so far used are, by their nature, not those that can, or

1 H. Antoniadis-Bibicou, Recherches sur les douanes a Byiance (Paris, I963), pp. I24-25, 152-53, etc.

2 An impression that would be difficult or impossible to prove, but which is generally accepted. Absolute numbers for western residents of the Empire are rare and probably exaggerated. Eustathius of Thessalonica reports that there were sixty thousand westerners in Constantinople in the second half of the twelfth century (Bonn edn, p. 394). The Historia Ducum Veneticorum (vi), claims that there were twenty thousand Venetians in Romania in I 70, of whom ten thousand were caught in Constantinople by Manuel's measures of 1171. See: ed. H. Simonsfeld, MGH, SS, xiv, p. 78.

3 R. S. Lopez in The Cambridge Economic History of Europe, ii, p. 309.

42 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

can be expected to, provide absolute or even comparative information.

The case that follows, forming the second part of this paper, is therefore constructed primarily from the evidence of the imperial coinage and the results of archaeological investigation, and only secondarily from the type of source utilized above. It rests upon the general assumption that, when backed by archaeology, the pattern of the coinage and the organizational system by which it was produced should, over an extended period and possibly at some relatively short chronological remove, bear some close relationship to the quality of contemporary economic life. In other words it should with all due qualification, allow of some comparative evaluation. A further, more particular, assumption is that the credit facilities available to the governments of the period were of an insufficiently large and flexible character to allow them to ride out a chronic and appreciable budgetary deficit without there being some visible and fairly immediate effect upon their coinage.'

The denominational pattern of the monetary system of the middle Byzantine period, from the eighth to the late eleventh century, compares unfavourably with those of the periods pre- ceding and succeeding it. Whereas that of the preceding period

1 Credit facilities on this scale demand the existence of an appreciable number of great merchants or bankers. There is no evidence for such a class in the Byzantine Empire, and in view of the continual governmental restric- tions placed upon private economic activity this is hardly surprising. When a Byzantine emperor needed large amounts of cash at short notice he resorted to the regalia, the Church or the magnates. But these were only short-term measures. When, at a later period, he floated loans, he went to foreign states-particularly Venice.

Psellus, in a laudatory description of Michael VII, describes him as knowledgeable in financial matters, but in listing them gives the distinct impression that this involved coinage and annual revenue and expenditure only, credit and loan being ignored completely. It is probably for this reason that Byzantine authors directly equate the physical emptiness of the treasury with imperial bankruptcy. See: Psellus, Michael VII (ii); ed. E. Renauld (Paris, 1928), ii, p. 173: Anna Comnena, 5 (i, ii); Bonn ed., i, pp. 225-27.

A recent study of the wills of two members of the Pacourianus fam- ily (Io9os), shows the kind of wealth available to magnates of the period. See: P. Tivchev and G. Changova-Petkova, 'Au sujet des relations f6odales dans les territoires bulgares sous la domination byzantine a la fin du XIe et pendant la premiere moitie du XIIe siecle', ByZantinobulgarica, ii, pp. io7- 125.

BYZANTIUM, IO8I--I204: AN ECONOMIC REAPPRAISAL 43

possessed an adequate fractional coinage in gold, the semissis (1) and tremissis (-), in addition to the unit, an admittedly ephemeral silver coinage, and a complex range of minor denominations in copper,' that of the middle period with rare exceptions consisted of three denominations only. These comprised the gold unit (nomisma), the silver twelfth (miliaresion) and the copper two hundred and eighty-eighth (follis), the last still involving a com- paratively large fraction.2 It was not until the monetary reforms carried out by Alexius I, in or about the year 1092, that a system with an appreciably more flexible spread of denominations came into circulation. This involved the gold unit (hyperpyron), the electrum third and billon forty-eighth (later the hundred and twentieth, both denominations being classed as trachea), and two copper denominations, the tetarteron and its half, of uncertain value but certainly less than the old follis.3

The organizational system which produced the coinage shows a parallel sequence of development. The intricate machinery of provincial mints that had marked the early period4 contracted even more than strictly required by the territorial losses that heralded its end. Despite the fact that parts of the outer Balkans and the whole of Asia Minor remained within the Empire, the mint of Thessalonica and those of Nicomedia, Cyzicus and Antioch ceased production during the first half of the seventh century, the Asian mints having apparently never recovered from the Persian invasions and Antioch having closed well before it

1 Basically consisting of the 40 nummia piece (the follis), and pieces of 20, Io, 5 and i nummus. Thessalonica and Alexandria struck on variant scales. A. R. Bellinger, Catalogue of the Byzantine Coins in the Dumbarton Oaks Collection and in the Whittemore Collection, i (Washington, D.C., 1966), pp. 3, 34, 64-65, I96-97, 264-65, 292-93; P. Grierson, Catalogue of the Byzantine Coins in the Dumbarton Oaks Collection and in the Whittemore Collection, ii (Washington, D.C., I969), pp. 8-32, 15I, 209, 242-43, 4I8-I9, 524, 574, 6II, 625, 647, 666, 674, 685.

2 From the reign of Leo III, the semissis and tremissis were struck on a nominal scale only; the half-follis was also struck occasionally (for instance under Theophilus, see p. 44, n. 2), but the generalization stands.

3 Hendy, op. cit., pp. 10-25. 4 The mints in more or less continuous operation (particularly for copper)

were: Constantinople, Thessalonica, Nicomedia, Cyzicus, Antioch, Alex- andria, Carthage, Sicily, Rome and Ravenna. Various others, mainly military, operated on a temporary basis. See: Bellinger, locc. citt; Grierson, locc. citt.

44 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

fell to the Arabs.1 Although a reform of the copper coinage under Theophilus (829-42) apparently required the temporary services of an additional mint, possibly Thessalonica,2 there is throughout no reason to believe that any other than Constantinople was nor- mally in operation.3 It was only during the reign of Constantine X (I059-67) that the mint of Thessalonica was reopened on a per- manent basis, and in about 1092 that a further provincial mint, probably located in central Greece, commenced production.4

Nothing absolute is known concerning the volume of the coin- age.5 The system of workshops (officinae) that had produced both gold and copper seems to have gradually contracted and finally disappeared during the eighth century,6 but this might conceivably be due to administrative rather than economic causes, as indeed might even be the case with the closing of provincial mints. It is noticeable, however, that their privy marks reappear on the precious-metal coinage of perhaps the tenth century and certainly the eleventh, reaching a climax in number and com- plexity during the twelfth. By this criterion, the production of

1 According to present information, Thessalonica closed in 630, Nico- media in 627, Cyzicus in 629, and Antioch in 6io. Grierson, op. cit., pp. 36, 37, 38, 4o.

2 This seems the obvious conclusion of the statistics given by D. M. Metcalf in 'The Reformed Folles of Theophilus: their Styles and Localiza- tion', American Numismatic Society Museum Notes, xiv (1968), pp. I32-33. Metcalf prefers to give his characteristic groups S and Z to separate mints in central or southern Greece rather than to Thessalonica, but this follows from his suggestion of a pattern of provincial mints that this author finds un- convincing. Metcalf's group A, which he assigns to Thessalonica, seems, in turn, to be a Constantinopolitan half-follis.

3 Despite attempts to prove the contrary. See in particular: D. M. Metcalf, 'The New Bronze Coinage of Theophilus and the Growth of the Balkan Themes', ANS Museum Notes, x (1962), pp. 8I-98-a first statement of the thesis later elaborated in the article quoted in note 2 above in the light of criticisms expressed by A. R. Bellinger in: 'Byzantine Notes', ANS Museum Notes, xiii (1967), pp. 136-41. Also, D. M. Metcalf, Coinage in the Balkans 820-z355 (Thessaloniki, I965), pp. 31-34; 'Provincial Issues among the

Byzantine Bronze Coinage of the IIth Century', Hamburger Beitrdge ~ur Numismatik, xv (I96I), pp. 31-35.

4 Hendy, op. cit., pp. 78, 98-IOI, 128-29.

5 A subject that has already provided considerable controversy. For the latest general statement, see P. Grierson, 'Byzantine Coinage as Source Material', Proceedings of the XIIIth International Congress of Byjantine Studies, Main Paper x, pp. 321-23.

6 Bellinger, 'Byzantine Notes', pp. 123-3I.

BYZANTIUM, IO8I-1204: AN ECONOMIC REAPPRAISAL 45

precious-metal (that is gold and electrum) tended to rise through- out the first half of the twelfth century, then levelled off without any hint of a radical decline at least until its last five years, when the evidence becomes uncertain.'

In the sophistication of its denominational pattern, the organ- izational system which produced it, and possibly even in its volume, the coinage of the twelfth century therefore represents a level equalled only by that of the sixth. It is certainly to be regarded as a more flexible and efficient instrument of exchange than the stable but rigid, sparse and in a word primitive system prevailing between the eighth and eleventh centuries.2 Although not entirely incompatible with an advanced and lively economy, the coinage of the middle period of Byzantine history betrays no hint of its existence.3

The evidence of the coinage as such is confirmed in two ways by that of archaeological excavation. The cities that have been excavated and from which the coin-finds are more or less known are, in Europe, Corinth and Athens, in Asia Sardis, Pergamum, Priene and Antioch.4 In all cases the pattern of finds is essentially the same, even if account is taken only of that denomination (the follis) which was common to the currency of the entire period from Anastasius I to Nicephorus III (491-Io8I). There are

present large numbers of coins of the emperors of the sixth and the first half of the seventh century, almost none of those of the

1 Hendy, op. cit., pp. I81-87, 315-I6. 2 The possible economic drawbacks to a coinage the predominant motive

behind which is stability have been pointed out by R. S. Lopez in 'The Dollar of the Middle Ages', Journal of Economic History, xi (I95 I), pp. 209-34.

3 The debasement of the gold coinage, which normally assumes such im- portance in accounts of the Byzantine coinage during the eleventh century, was in fact a temporary phenomenon probably caused more by imperial irresponsibility and eventual military defeat and its consequences rather than by long-term economic trends. If the last had ever existed they had apparently ceased to be operative by the year 1092. See Hendy, op. cit., pp. 316-19.

4 K. M. Edwards, Corinth, vi, Coins (Cambridge, Mass., 1933), pp. 121-50; M. Thompson, The Athenian Agora, ii, Coins from the Roman through the Venetian Period (Princeton, 1954), pp. 67-75; H. W. Bell, Sardis, xi(i), Coins (Leiden, I9I6), pp. 76-108; S. McA. Mosser, A Bibliography of Byzantine Coin Hoards, ANS Numismatic Notes and Monographs, lxvii (New York, I935), pp. 64-65, 70; D. B. Waage, Antioch-on-the-Orontes, iv(ii), Greek, Roman, Byzantine and Crusaders' Coins (Princeton, 1952), pp. 148-68.

TRANS. 5TH S.-VOL. 20-D

46 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

period from the second half of the seventh down to (and includ- ing) the first half of the ninth century, and then at first slowly, later rapidly, increasing numbers to the end of the eleventh cen- tury.1 The representation from the twelfth century is, in Europe, quite overwhelming, in Asia less strong although it picks up again during the second half of the century and continues into the thirteenth. This difference may be due to reasons of coin circulation as well as to political events.2

There is a small amount of additional evidence confirming the pattern common to these major series. Such are the few Byzantine coins from the Troy excavations-and this despite the fact that, although by now a mere village, it lay near the major north- south sea route.3 Metcalf has noticed the same for surface finds in south-central Anatolia.4

At Corinth and Athens the numismatic evidence can be checked by reference to urban construction and population. At Corinth the depressed remnant of the late Roman and early Byzantine city stagnated throughout the seventh and eighth centuries. An appreciable but still slow recovery took place from the beginning of the ninth century until, to quote Scranton: 'Apparently construction accelerated around the middle of the eleventh century, reaching a climax and almost filling the whole Agora area around the middle of the twelfth century.... As we have frequently observed, the developed form of the mediaeval community at Corinth was that of the second half of the twelfth century.'5 This expansion was industrial and entrepreneurial as well as residential. The same sequence of development seems

1 The series from Antioch is, it is true, anomalous, since the city was in Arab hands during the seventh to tenth centuries. It is noticeable, however, that although it fell to the Arabs shortly after the battle of Yarmuk in 636, Byzantine coins continue to appear up to and including the reign of Constans II (641-68) and only then cease. See Waage, op. cit., pp. I64-66.

2 The dominant element of the circulating medium in Greece seems to have been the low-value copper tetarteron and its half; that in Asia Minor the higher-value billon trachy which would, because of its value, tend to have been lost less. See Hendy, op. cit., p. 3II.

3 A. R. Bellinger, Troy, Supplementary Monograph ii, the Coins (Princeton /Cincinnati, I96I), pp. I8I-82.

4D. M. Metcalf, 'How Extensive was the Issue of Folles during the Years 775-820?' By[antion, xxxvii (1967), p. 306.

5 R. L. Scranton, Corinth, xvi, Mediaeval Architecture in the Central Area of Corinth (Princeton, I957), pp. 53, 57.

BYZANTIUM, IO8I-1204: AN ECONOMIC REAPPRAISAL 47

clear at Athens,1 and Condurachi, Barnea and Diaconu have reported a significant increase in the intensity of occupation at sites on the lower Danube from the reign of Alexius I onwards, accompanied by the more constant appearance of imported objects.2 Mango and Hawkins have noticed the increasingly fre- quent foundation, endowment and decoration of religious com- munities and churches in Cyprus during the course of the twelfth century.3 This they attribute to the growing political importance of the island and a consequent enhancement of imperial interest, but while such an explanation may contain an element of truth the phenomenon is better understood against a background of economic and even demographic expansion particularly since the majority of cases involve not public and imperial, but private and aristocratic benefactions. A brief survey also suggests that the phenomenon might not be confined to Cyprus4 and brings to mind interesting western parallels.

The consistency of the evidence provided by the coinage, and by coin-finds on archaeological sites, is impressive, and it may be suggested that it is now time to accept the fact that this particular pattern of occurrence is likely to be more or less standard, with the possible exception of series from such cities as Constantinople and Thessalonica.5 The crucial point is, however, that unless Corinth and Athens are completely atypical what produces this pattern are parallel developments in population and construction.

1 For instance, T. L. Shear, 'The Campaign of 1936', Hesperia, vi (1937), p. 342.

2 E. Condurachi, I. Barnea, P. Diaconu, 'Nouvelles recherches sur le Limes byzantin du Bas-Danube aux Xe-XIe siecles', Proceedings of the XIIIth International Congress of Byzantine Studies, Main Paper vi, p. I93.

3 C. Mango and E. J. W. Hawkins, 'The Hermitage of St Neophytos and its Wall Paintings', Dumbarton Oaks Papers, xx (1966), pp. 204-6.

4 It is once more difficult to attempt a comparison with other periods, but an article by C. Delvoye ('L'architecture byzantine au XIe siecle', Pro- ceedings of the XIIIth International Congress of Byzantine Studies, pp. 225-34) seems to suggest a peak of construction at about this time.

5 The coin series from excavations in and around the hippodrome at Constantinople during the years 1927/28 show less consistent signs of this pattern, although traces of the decline between the early eighth and mid- ninth centuries still occasionally occur. See A. H. M. Jones in Preliminary Report upon the Excavations Carried Out in the Hippodrome of Constantinople in z.927 (London, 1928), pp. 46-50, and B. Gray in Second Report upon the Excavations Carried Out In and Near the Hippodrome of Constantinople in 1928 (London, I929), p. 50.

48 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

Two recently published articles add weight to this conclusion. Tivchev, in a short study of the urban history of the Empire between Io8I and 1204, based mainly on the documentary evi- dence of Benjamin of Tudela, Edrisi, and contemporary Byzantine chroniclers, views the period as one of expansion in urban popula- tion, industry and trade-particularly in the European provinces: a position accepted in principle by Charanis.1 Finally, Goitein has published the text of a letter from a Jewish immigrant written at Seleucia in Isauria (then, in 1137, an area of political and economic importance) to a relative in Cairo. In the course of the letter the immigrant describes his present material pros- perity and encourages his relative to follow his own example, and that of eleven other Jews known to them, and to emigrate to the Empire. Goitein comments: 'This reflects the state of relative security and prosperity enjoyed by the Byzantine Empire at that time, over sixty years after the Seljuk invasion'.2

III

While the continued existence of cities, industry and trade during the seventh to tenth centuries cannot be doubted, the physical evidence for the quality of contemporary urban life is consistently unfavourable to the supposition that it was com- parable to that of the sixth or the eleventh and twelfth centuries. Pergamum and Priene were ports, and Sardis had formerly been one of the major cities of the western river-valley and road system of Asia Minor. Yet their evidence differs in no great respect from that of Corinth and Athens. Urban population was small; it did not have an efficient means of exchange at its disposal, and such small change as was provided was evidently little used, since it was little lost. There can have been little industry and therefore trade cannot have been on any large scale, except possibly where basic materials and foodstuffs were concerned.3 It is submitted

1 P. Tivchev, 'Sur les cites byzantines au XIe-XIIe siecles', Byjantino- bulgarica, i (I962), pp. I45-82; Charanis, 'Observations on the Demography of the Byzantine Empire', p. 460.

2 S. D. Goitein, 'A letter from Seleucia (Cilicia) dated 21 July 1137', Speculum, xxxix (I964), pp. 298-303.

3 Cf E. Frances, 'L'empereur Nicephore ier et le commerce maritime

BYZANTIUM, IO8I-I204: AN ECONOMIC REAPPRAISAL 49

that the material used to produce this conclusion is of a kind qualitatively and quantitatively more reliable than that used to produce a contrary one, largely based on the occasional docu- mentary reference and employing unverifiable figures likely to lend it a spurious air of exactitude. Comparison with contempor- ary western material shows that the use of monetary terms in legal documents is not necessarily a reliable indication of the existence of a monetary economy of any depth, and an acquaint- ance with the monetary and taxation policies of the late Roman and early Byzantine Empires shows the same to be potentially true of an apparent abundance of gold coin.'

Those who would assert the existence of a flourishing industry and trade as a corollary to a vital urban life at this period have not only the lack of physical evidence to contend with, but where external trade is concerned, a further and more general considera- tion. This is that the Byzantines have to be shown to be capable of producing appropriate quantities of goods; the means and opportunity of shipping them have to be shown to have existed; and, finally, other communities have to be shown to be capable of paying for and absorbing them. The evidence for this is singularly unpromising.

All that is known of contemporary societies in the west, the inner Balkans, and Russia from the seventh to tenth centuries shows them to have been at a low level of economic development. In this context it should be remembered that as late as the reign of Michael IV (I034-41) a Byzantine attempt to switch from taxation in kind to taxation in coin in the recently reconquered Bulgarian themes had provoked a serious revolt.2 The evidence for large-scale trade with and within the west is tenuous, to say the least,3 and such exports as there were from these three regions

byzantin', By{antinoslavica, xxvii (I966), pp. 41-47; J. L. Teall, 'The Grain Supply of the Byzantine Empire, 330-1025', Dumbarton Oaks Papers, xiii (I959), pp. 89-I39.

1 C. M. Cipolla, Money, Prices and Civilization in the Mediterranean World (Princeton/Cincinnati, I956), pp. 3-12; J. P. C. Kent, 'Gold Coinage in the Later Roman Empire', in Essays in Roman Coinage Presented to Harold Mattingly (Oxford, I956), pp. 190-204.

2 Cedrenus; Bonn edn, ii, p. 530. 3 P. Grierson, 'Commerce in the Dark Ages: a Critique of the Evidence',

Trans. Roy. Hist. Soc., Fifth Series, ix (I959), pp. 123-40.

50 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

were almost entirely of a primary nature.l In default of large amounts of money or bullion to make up the difference they would have sufficed to purchase only a limited amount of manufactured, luxury goods that are known to have been the goal of the exchange,2 and the large-scale sale of which might have stimulated Byzantine industry. Strict governmental regulation of the quality, quantity and type of Byzantine exports is unlikely to have made the process any easier.3 Although trade with the more advanced Arab countries of the east and south might have provided the required industrial stimulus and some degree of such trade can be shown to have existed, the main trade-axis of the Arab world seems, on good evidence, to have been aligned on an east-west basis rather than a north-south one.4

It is the eleventh and twelfth centuries, on the contrary, that represent the apogee of Byzantine mercantile develop- ment, the fragmented economies of the thirteenth-despite a modest and mainly agricultural prosperity in Asia Minor5- being too overshadowed by the colossal disaster of the Fourth Crusade.

This development was doubtless the product of many and

1 For the Russian trade see: A. A. Vasiliev, 'Economic Relations between Byzantium and Old Russia', Journal of Economic and Business History, iv (I931-32), pp. 314-34. For the western see: Urkunden {ur alteren Handels- und Staatsgeschichte der Republik Venedig mit besonderer Beriehung auf Byranr und die Levante, eds. G. Tafel and G. Thomas, i (Vienna, i856), no. xiii, pp. 19-25, no. xiv, pp. 26-33, no. xvii, pp. 36-39 (for the apparent balance of trade); Ibn Khordadhbeh, trans. cit., pp. II4-I6; Goitein, A Mediterranean Society, i, pp. 46-47. For the Bulgarian see: the Book of the Prefect (9 [vi], eds. J. Zepos and P. Zepos in lus Graecoromanum, ii [Athens, I93I]); infra, p. I, n. 2.

2 Vasiliev, 'Economic Relations', p. 325. 3 Prohibition of the export of gold is found in both the Codex Justinianus

(4, 63, 2) and in the Basilika (56, I, 20), of precious metals and stones in the Book of the Prefect (2[iv]). Various prohibitions and restrictions on the export of silk are to be found in the Book of the Prefect (4[i], [viii]; 6[xvi]; 8[iii], [v]) and that they were enforced is shown by the experience of Liut- prand of Cremona-J. Becker, Die Werke Liudprands von Cremona (Leipzig, I915), pp. 204-06. Russian merchants were, after 945, allowed to purchase only fifty nomismata worth of silk per person: The Russian Primary Chronicle, Laurentian Text, trans. S. H. Cross, and 0. P. Sherbowitz-Wetzor (Cambridge, Mass., 1953), p. 75-

4 Goitein, A Mediterranean Society, i, pp. 211-14. 5

Nicephorus Gregoras 2, vi; Bonn ed. i, pp. 42-43.

BYZANTIUM, I08-1I204: AN ECONOMIC REAPPRAISAL 51

complex factors, both external and internal. Several of those that had militated against the evolution of an advanced economy at an earlier period were no longer operative. It seems clear that western Europe was undergoing a general and appreciable rise in popula- tion and living-standards: it should therefore have become more capable of absorbing imports and of paying for them with pro- ducts other than raw materials. It has even been suggested that a favourable balance of trade with the Muslim states of North Africa resulted in its acquisition of a surplus in gold which would have permitted it to offset an unfavourable one with the Empire and the east.' The establishment of the Crusader states in the Levant-itself possibly the expression of an increase in popula- tion-would have necessitated a more frequent sea-traffic that could hardly have remained divorced from commerce even if not originally based on it.

The fact that the Empire seems to have been experiencing at least an urban expansion of its own would have meant that many of the prerequisites for a period of rapid growth in international trade were present. This would have been potentially beneficial to east and west, whoever provided the bulk of shipping for transport. In comparison, it may be suggested that the loss of central Asia Minor was something of an irrelevance, particularly since the inner Balkans were now available as an alternative source of agricultural produce. Bulgaria, after a century and a half of Byzantine occupation, seems to have been so firmly integrated into the Byzantine system that the Asenid revolt of the ii 8os had little immediate economic effect: after all, the natural outlets for its products would have remained the ports of the Black Sea and northern Aegean.2

If one may express two general criticisms of scholarship relat- ing to the economic history of the Byzantine state, they are these: that too great an emphasis has been laid upon the political and economic stability that is said to have marked the ninth and tenth centuries; and that there has been a tendency to assume that,

1 Lopez in The Cambridge Economic History ofEurope, ii, p. 309. 2 Trade in agricultural products between the Black Sea ports and the

capital seems to have been of considerable importance as late as the four- teenth century. Its results may be seen in the hoards of late Byzantine gold hyperpyra found in the region and its hinterland. For the connexion see: T. Gerassimov, 'Les hyperperes d'Andronic II et d'Andronic III et leur circulation en Bulgarie', By{antinobulgarica, i, pp. 213-36.

52 TRANSACTIONS OF THE ROYAL HISTORICAL SOCIETY

because the western domination which characterizes the thir- teenth and particularly the fourteenth centuries was so blatantly detrimental to the east, both the balance and direction of the relationship must have already been present during the eleventh and twelfth centuries. This combination has, in part at least, been responsible for reversing the actual sequence of economic devel- opment for the earlier period; which is that rather than declining away from the ninth and tenth centuries, Byzantine economic life was expanding rapidly throughout the eleventh and twelfth centuries; and that despite the beginnings of political reverses under the Angeli (and mercenaries and magnates notwith- standing), this expansion may have only come to an end with the Fourth Crusade.

Fitswilliam Museum, Cambridge