Digitized by the Internet Archive in 2007 with funding from IVIicrosoft Corporation http://www.archive.org/details/constitutionoflaOOburyrich THE CONSTITUTION OF THE LATER ROMAN EMPIRE CAMBRIDGE UNIVERSITY PRESS EottiJon: FETTER LANE, E.G. 0. F. CLAY, Manager d^trinburgfj: 100, PRINCES STREET 38erltn: A. ASHER AND CO. Eeipjts: E. A. BROCKHAUS iplefa Sorft; G. P. PUTNAM'S SONS aSomftag airtj (JTsIcutta: MACMILLAN AND CO., Ltd. All rights reserved THE CONSTITUTION OF THE LATER ROMAN EMPIRE CREIGHTON MEMORIAL LECTURE DELIVERED AT UNIVERSITY COLLEGE, LONDON 12 NOVEMBER 1909 BY J. B. BURY REGIUS PROFESSOR OF MODERN HISTORY IN THE UNIVERSITY OF CAMBRIDGE Cambridge : at the University Press 1910 ^f^ PKINTED BY JOHN CLAY, M.A. AT THE UNIVBB8ITY PRESS ,♦,' THE CONSTITUTION OF THE LATER ROMAN EMPIRE The forms of government which are commonly classified as absolute monarchies have not received the same attention or been so carefully analysed as those forms which are known as republics and constitutional monarchies. There is a considerable litera- ture on absolute monarchy considered theoretically, in connexion with the question of Divine Right, but the actual examples which history offers of this kind of govern- ment have not been the subject of a detailed comparative study. Montesquieu, for in- stance, treats them indiscriminately as despotisms. Probably the reason lies in the apparent simplicity of a constitution, by which the supreme power is exclusively vested in one man. When we say that the monarch's will is supreme, we may seem to 261382 THE CONSTITUTION OF say all there is to be said. The Later Roman Empire is an example of absolute monarchy, and I propose to shew that so far as it is concerned there is a good deal more to be said. The term absolute monarchy is applied in contradistinction to limited or constitu- tional monarchy. I understand the former to mean that the whole legislative, judicial, and executive powers of the state are vested j in the monarch, and there is no other in-;^ dependent and concurrent authority^. The latter means that besides the so-called monarch there are other political bodies which possess an independent and effective authority of their own, and share in the sovran power. These terms, absolute and constitutional monarchy, are unsatisfactory, from a logical point of view. For they group together these two forms of government as subdivisions of the class monarchy, implying or suggesting that they have much more real afl&aity to one another than either has to other constitutions. This is evidently untrue: THE LATER ROMAN EMPIRE 3 a constitutional monarchy is far more closely allied to a republic like France than to an absolute monarchy like Russia. The English constitution, for instance, in which legislation is effected by the consent of three in- dependent organs, the Crown, the Lords, and the Commons, might be described more correctly as a triarchy than as a monarchy ; and it seems to be unfortunate that monarchy should have come to be used, quite un- necessarily, as a synonym for kingship. "Limited monarchy," as Austin said long ago, "is not monarchy^"; monarchy properly so-called is, simply and solely, absolute monarchy. We have however an alternative term, "autocracy," which involves no am- biguities, and might, I venture to think, be advantageously adopted as the technical term for this form of government in con- stitutional discussions. And "autocracy" has a special advantage over "absolute monarchy." Autocracies are not all alike, in respect to the power actually exercised by the autocrat. Although not limited by any bodies pos- 1—2 THE CONSTITUTION OF sessing an independent authority, he may be limited effectually in other ways. Now we can properly speak of more or less limited autocracies, whereas it is an impropriety of language to speak of more or less absolute monarchies, as "absolute" admits of no degrees. Originally, and during the first three centuries of its existence, the Roman Empire was theoretically a republic. The Senate co-existed with the Emperor, as a body invested with an authority independent of Jiis ; but the functions which it exercised by virtue of that authority were surrendered one by one; it became more and more dependent on him ; and by the end of the third century the fiction of a second power in the state was dropped altogether, although the Senate was not abolished^. From that time forward, under the system established by Diocletian and Constantine, until the fall of the Empire in the fifteenth century, the government was simply and undisguisedly an autocracy. THE LATER ROMAN EMPIRE 5 Now one broad distinction between auto- cracies may be found in the mode of '^ accession to the throne. The sovranty may be hereditary or it may be elective. If it ^ is elective, the sovranty is derived from the electors who, when the throne is vacant, exercise an independent and sovran authority in electing a new monarch. If it is hereditary, if the right of the autocrat depends entirely and indefeasibly on his birth, then we may say that his sovranty is underived ; the suc- cession is automatic, and there is no moment at which any other person or persons than the monarch can perform an act of sovran authority such as is implied in the election of a sovran. This difference may involve, as we shall see, important consequences. In the case of the Roman Empire, the Imperial dignity continued to be elective, as it had been from the beginning, and the method of election remained the same. When the throne was vacant a new Emperor was chosen by the Senate and the army. The initiative might be taken either by the Senate THE CONSTITUTION OF or by the army, and both methods were ^ recognised as equally valid. It was of course ^ only a portion of the army that actually chose an Emperor, — for instance, if the choice were made in Constantinople, the guard regiments ; but such a portion was regarded as for this purpose representing all the troops which were scattered over the Empire. The appointment did not take the formal \ shape of what we commonly understand by election. If the soldiers took the initiative, they simply proclaimed the man they wanted. If the choice was made by the Senate, the procedure might be more deliberate, but there seems to have been no formal casting ? of votes, and the essential act was the proclamation^. It sufficed that one of these r bodies should proclaim an Emperor to establish his title to the sovranty; it only remained for the other body to concur ; and the inauguration was formally completed when the people of Constantinople had also acclaimed him in the Hippodrome — a^ formality always observed and reminiscent THE LATER ROMAN EMPIRE 7 of the fact that the inhabitants of the new capital of Constantine had succeeded to the position of the old popidiis Romanus^. The part which the Senate played in the appointment of an Emperor, whether by choosing him or by ratifying the choice of the army, is constitutionally important. The Senate or Synkletos of New Kome was a very different body from the old Senatus Komanus. It was a small council consisting of persons who belonged to it by virtue of administrative offices to which they were appointed by the Emperor. In fact, the old Senate had coalesced with the Consistorium or Imperial council, and in consequence the new Senate had a double aspect. So long as there was a reigning Emperor, it acted as consistorium or advisory council of the sovran, but when there was an interval between two reigns, it resumed the independent authority which had lain in abeyance and performed functions which it had inherited from the early Senate. But it was not only when the throne was vacant that it could perform such functions. THE CONSTITUTION OF The right of election might be exercised by the Senate and the army at any time. It • was a principle of state-law in the Early Empire that the people which made the Emperor could also unmake him, and this principle continued in force under the autocracy. There was no formal process of deposing a sovran, but the members of the community had the means of dethroning him, if his government failed to give satisfaction, \ by proclaiming a new Emperor ; and if anyone so proclaimed obtained sufficient support from the army, Senate, and people, the old Emperor was compelled to vacate the throne, retiring into a monastery, losing his eyesight, or suffering death, according to the circumstances of the situation or the temper of his supplanter; while the new Emperor was regarded as the legitimate monarch from the day on which he was proclaimed; the proclamation was taken as the legal expression of the general will. If he had not a sufficient following to render the proclamation effective and was sup- THE LATER ROMAN EMPIRE 9 pressed, he was treated as a rebel; but during ^ the struggle and before the catastrophe, the fact that a portion of the army had proclaimed him gave him a presumptive constitutional status, which the event might either confirm or annul. The method of deposition was in fact revolution, and we are accustomed to regard revolution as something essentially unconstitutional, an appeal from law to force; but under the Imperial system, it was not unconstitutional; the government was, to use an expression of Mommsen, "an auto- cracy tempered by the legal right of revolu- tion." Thus the sovranty of the Koman autocrat was delegated to him by the community, as represented by the Senate, and the army, and, we may add, the people of Constantinople^. The symbol of the sovranty thus delegated was the diadem, which was definitely intro-\ duced by Constantino. The Emperor wore other insignia, such as the purple robe and the red boots, but the diadem was pre- eminently the symbol and expression of the 10 THE CONSTITUTION OF autocracy. The dress only represented the Imperator or commander-in-chief of the army, and no formalities were connected with its assumption. It was otherwise with the crown, which in the Persian Kingdom, from which it was borrowed, was placed on the king's head by the High-priest of the Magian religion. In theory, the Imperial crown should be imposed by a representative of those who conferred the sovran authority ' which it symbolized. And in the fourth century we find the Prefect, Sallustius Secundus, crowning Valentinian I, in whose election he had taken the most prominent part. But the Emperors seem to have felt some hesitation in thus receiving the diadem ^ from the hands of a subject; and the selection of one magnate for this high office of con- ferring the symbol of sovranty was likely to cause enmity and jealousy. Yet a formality was considered necessary. In the fifth century, the difficulty was overcome in a clever and tactful way. The duty of corona- tion was assigned to the Patriarch of Con- THE LATER ROMAN EMPIRE 11 stantinople. In discharging this office, the Patriarch was not envied by the secular magnates because he could not be their rival, and his ecclesiastical position relieved the Emperor from all embarrassment in receiving the diadem from a subject. There is some evidence, though it is not above suspicion, that this plan was adopted at the coronation of Marcian in a.d. 450, but it seems certain that his successor Leo was crowned by the Patriarch in a.d. 457. Henceforward this was the regular practice. In the thirteenth century we find Theodore II postponing his coronation until the Patriarchal throne, which happened to be vacant, was filled. But although it was the regular and desirable form of coronation, it was never regarded as indispensable for the autocrat's legitimate inauguration. The last of the East Roman Emperors, Constantino Palaeologus, was not crowned by the Patriarch; he was crowned by a layman '^. This fact that coronation by the Patriarch was not constitutionally neces- sary, though it was the usual custom, is 12 THE CONSTITUTION OF significant. For it shows that the Patriarch, in performing the ceremony, was not repre- senting the Church. It is possible that the idea of committing the office to him was suggested by the Persian coronations which were performed by the High-priest, but the significance was not the same. The chief of the Magians acted as the representative of the Persian religion, the Patriarch acted as the representative of the State ^. For if he had specially represented the Church, it is clear that his co-operation could never have been dispensed with. In other words, no new constitutional theory or constitutional requirement was introduced by the assign- ment of the privilege of crowning Emperors to the Patriarch. It did not mean that the consent of the Church was formally necessary to the inauguration of the sovran. I will make this point still more evident presently in connexion with another im- portant feature of the constitution to which we now come. If you look down the roll of Emperors, you will find that only a minority THE LATER ROMAN EMPIRE 13 of them were actually elected in the ways I have described. In most cases, when an Emperor died, the throne was not vacant, for generally he had a younger colleague, who had already been invested with the Imperial dignity, so that no new election was necessary. This practice^ by which a reign- ing Emperor could appoint his successor modified the elective principle. The Emperor used to devolve the succession upon his son, if he had one; so that son constantly suc- ceeded father, and the history of the Roman Empire is marked by a series of hereditary dynasties. The constitution thus combined the elective and the hereditary principles; a device was found for securing the advan- tages of hereditary succession, and obviating its disadvantages by preserving the principle of election. The chief advantage of here- ditary monarchy is that it avoids the danger of domestic troubles and civil war which are likely to occur when the throne is elective, and there are two rival candidates. Its chief disadvantage is that the supreme power in 14 THE CONSTITUTION OF the State will inevitably devolve sometimes upon a weak and incapable ruler. The result of the mixture of the two principles, the dynastic and the elective, was that there were far fewer incapable sovrans than if the dynastic succession had been exclusively valid, and fewer struggles for power than if every change of ruler had meant an election. It would be interesting to trace, if we had the material, how the inhabitants of the Empire became more and more attached to the idea of legitimacy — the idea that the children of an Emperor had a constitutional right to the supreme power. We can see at least that this feeling grew very strong under the long rule of the Macedonian dynasty; it is illustrated by the political role which the Empress Zoe, an utterly incompetent and depraved old woman, was allowed to play because she was the daughter of Con- stantine VIII. But the fact remained that although a father invariably raised his eldest son, and sometimes younger sons too, to the rank of Augustus, the son became Emperor THE LATER ROMAN EMPIRE 15 by virtue of his father's will and not by virtue of his birth. The Emperor was not in any way bound to devolve the succession upon his son^^. Now what I ask you to observe is that when a reigning sovran created a second Emperor, whether his son or anyone else, there was no election. The Senate, the army, and the people expressed their joy and satisfaction, in the ceremonies which attended the creation, but the creation was entirely the act of the Emperor. The constitutional significance is evident. The autocratic powers conferred upon an Emperor by his election included the right of devolving the Imperial dignity upon others. It was part of his sovranty to be able to create a colleague who was potentially another sovran. This difference between the appointment of an Emperor when the throne is vacant and the appointment of an Emperor as colleague when the throne is occupied is clearly and significantly expressed by the difference between the coronation acts in the two cases. In the former case the act is performed by 16 THE CONSTITUTION OF a representative of the electors, almost always the Patriarch ; in the latter case it is regu- larly performed by the reigning Emperor. It is he who, possessing the undivided sovranty, confers the Imperial dignity and therefore with his own hands delivers its symbol. Some- times indeed he commits the office of corona- tion to the Patriarch, but the Patriarch is then acting simply as his delegate ^^. This difference is a confirmation of the view that the Patriarch, in discharging the duty of coronation, acts as a representative of the electors, and not of the Church. For if the coronation had been conceived as a religious act, it must have been performed in the same way, in all cases, by the chief minister of the Church. But now you may ask, is the term auto- cracy or the term monarchy strictly applic- able to the Empire? Monarchy and autocracy mean the sovran rule of one man alone, but, as we have just seen, the Emperor generally had a colleague. Both in the early and in the later Empire, there were constantly two THE LATER ROMAN EMPIRE 17 Emperors, sometimes more. In the tenth cen- tury, for instance, in the reign of Eomanus I, there were as many as five — each of them an Augustus, each a Basileus^^. This practice is derived from the original collegial character of the proconsular Imperium and the tri- bunician power, on which Augustus based his authority. But, although the Koman Im- perium or Basileia was collegial, the sovranty was not divided. When there were two Emperors only one exercised the sovran power and governed the State ; his colleague was subordinate, and simply enjoyed the dignity and the expectation of succession. Though his name appeared in legislative acts and his effigy on coins, and though he shared in all the Imperial honours, he was a sleeping partner. With one exception, which I will notice presently, the only cases of Imperial colleagues exercising concurrent sovranty were in the period from Diocletian to the death of Julius Nepos, when the Empire was territorially divided. Diocletian and Maximian, for instance, the sons of Constan- B. 2 18 THE CONSTITUTION OF tine, Arcadius and Honorius, were severally monarchs in their own dominions. But ex- cept in the ease of territorial division, the supreme power was exercised by one man, and monarchy is therefore a right descrip- tion of the constitution. In the reign of Constantine IV, the soldiers demanded that the Emperor should crown his two brothers. " We believe in the Trinity," they cried, " and we would have three Emperors." But this must not be interpreted as a demand that each member of the desired Imperial trinity should exercise sovran authority. Such a joint sovranty was never tried except in one case, and a clear distinction was drawn between the Basileus who governed and the Basileus who did not govern. The exceptional case was the peculiar one of two Empresses, who ruled conjointly for a short time in the eleventh century. I will mention this case again, in a few minutes, when I come to speak of the position of Empresses. And here I must dwell for a moment on the name Basileus and another Greek name THE LATER ROMAN EMPIRE 19 Autokrator, which were employed to designate the Emperor, tn the early Empire, Basileus was used in the East and especially in Egypt, where Augustus was regarded as the suc- cessor of the Ptolemies, but it was not used officially by the Emperors; it was not the Greek for Imperator. The Greek word adopted to translate Imperator was Auto- krator, and this is the term always used in ]/ Imperial Greek inscriptions. By the fourth century Basileus had come into universal use in the Greek-speaking parts of the Empire; it was the regular term used by Greek writers ; but it was not yet accepted as an official title. Nor was it adopted officially till the seventh century in the reign of Heraclius. It has been pointed out by Brdhier ^^ that the earliest official act in which an Emperor entitles himself Basileus is a law ^ of Heraclius of the year 629. In the earlier diplomas of his reign he uses the old tra- ditional form Autokrator. Brdhier, however, has failed to see the reason of this change of style, but the significant date a.d. 629 sup- 2—2 20 THE CONSTITUTION OF plies the explanation. In that year Heraelius completed the conquest of Persia. Now, the Persian king was the only foreign monarch to whom the Roman Emperors conceded the title Basilens; except the Abyssinian king, who hardly counted. So long as there was a great independent Basilens outside the Roman Empire, the Emperors refrained from adopting a title which would be shared by another monarch. But as soon as that monarch was reduced to the condition of a dependent vassal and there was no longer a concurrence, the Emperor signified the event by assuming officially the title which had for several centuries been applied to him un- officially. The Empire was extremely con- servative in forms and usages ; changes were slow in official documents, they were slower still in the coinage. It is not till more than a century later that Basilens begins to be adopted by the mint. By this change Basilens became the official equivalent of Imperator ; it took the place of Autokrator ; and it was now possible for Autokrator to come into its THE LATER ROMAN EMPIRE 21 own and express its full etymological signifi- cance. Thus we find a strongly marked tendency in later times to apply the term specially to the Basileus who was the actual ruler. Though he and his colleague might be acclaimed jointly as Autokrators; yet Autokrator is distinctly used to express the plenitude of despotic power which was exercised by the senior Emperor alone ^^. Thus we may say that in early times Basileus was the pregnant title which expressed that full monarchical authority which the system of Augustus aimed at disguising, and Auto- krator was simply the equivalent of the republican title Imperator; while in later times the roles of the two titles were re- versed, and Autokrator became the pregnant title, expressing the fulness of authority which the familiar Basileus no longer em- phasized. Before we leave this part of our subject, a word must be said about the rights of women to exercise autocracy. From the foundation of the Empire the title of Augusta had been 22 THE CONSTITUTION OF conferred on the wives of Emperors, and we find in early times the mothers of minors, like Agrippina and Julia Domna, exercising political power. But this power was always exercised in the name of their sons. At the beginning of the fifth century the Augusta Pulcheria presides over the government which acted for her brother Theodosius II while he was a minor. On his death without children, it is recognised that although she cannot govern alone, she nevertheless has a right to have a voice in the election of a new Emperor, and the situation is met by her nominal marriage with Marcian. Simi- larly, forty years later, when Zeno dies without a son, his wife, the Augusta Ariadne, has, by general consent, the decisive voice in selecting her husband's successor; her choice falls on Anastasius, and he is elected. But it is not she who confers the Imperial authority on Anastasius, it is the Senate and army, who elect him, in accordance with her wishes. In the following century, the political importance of Empresses is augmented by the exceptional THE LATER ROMAN EMPIRE 23 positions occupied by Theodora the consort of Justinian, and Sophia the consort of Justin II. But so far although an Empress may- act as regent for a minor^^, may intervene in an Imperial election, may receive honours suggesting that she is her husband's colleague rather than consort, she never exercises inde- pendent sovran power, she is never, in the later sense of the word, an Autokrator. Passing on to the close of the eighth century, we come to the Empress Irene, the Athenian lady who is famous as the first restorer of Image-worship. When her husband died, her son Constantino was too young to rule, and she governed in the same way as Pulcheria had governed for Theodosius. When Con- stantino was old enough to govern himself, Irene was unwilling to retire into the back- ground, and although the son succeeded in holding the power in his own hands for some years, the mother was continually in- triguing against him. The struggle ended in her triumph. She caused her son to be blinded, and for five years she reigned alone 24 THE CONSTITUTION OF with full sovran powers as Autokrator. This was a considerable constitutional innovation, and the official style of her diplomas illus- trates, in an interesting] way, that it was felt as such. She was, of course, always spoken of as the Empress, but in her official acts she is styled not "Irene the Empress" but "Irene the Emperor" {BasileusY^, It was felt that only an Emperor could legislate, and so the legal fiction of her masculinity was adopted. It was said in Western Europe, for the purpose of justifying the Imperial claim of Charles the Great, that the sovranty of the Empire could not devolve on a woman, and that Irene's tenure of power was really an interregnum; but the Byzantines never ad- mitted this constitutional doctrine. Never- theless they had a strong objection to the regime of women, except in the capacity of regents, and the precedent established by Irene was repeated only in the case of Zoe and Theodora, the two nieces of Basil II. We find each of these ladies exercising the sovran authority alone for brief periods, and THE LATER ROMAN EMPIRE 25 we also find them ruling together. This is the instance, which I mentioned already, of the experiment of government by two auto- crats. Their joint rule might have been pro- tracted, if they had been in harmony, but Zoe was extremely jealous of Theodora, and in order to oust her she took a husband, who immediatelyassumed the autocratic authority, and Zoe fell back into the subordinate position of a consort. We may now pass to the consideration of the nature and amplitude of the Imperial supremacy. The act of proclamation con- ferred his sovran powers upon the Emperor. In early days the Imperial powers were de- fined explicitly by a law, the lex de imperio. We have the text of the law which was passed for Vespasian. But the practice of passing it anew on the accession of a new Emperor was discontinued, and under the autocracy, when all the legislative, judicial and executive powers were vested in the autocrat, there was no reason to define what those powers were. In the sixth century however, in the THE CONSTITUTION OF legislation of Justinian, it is recognised that by the lex de imperio the people transferred its sovranty to the Emperor. In the eighth century we may be pretty sure that no one from the Emperor downwards had ever heard of the lex de imperio^'^. But although there was no constitution of this kind defining or limiting the monarch's functions, I will pro- ceed to shew that his power, legally unlimited, was subject to limitations which must be described as constitutional. For his legislative and administrative acts, the monarch was responsible to none, except to Heaven ; there was no organ in the state that had a right to control him ; so that his government answers to our definition of auto- cracy. But when the monarch is appointed by any body or bodies in the state, the electors can impose conditions on him at the time of election, and thus there is the possibility of limiting his power. In other words, an elective autocracy, like the Roman Empire, is liable to the imposition of limitations. The case of the Emperor Anastasius I is in THE LATER ROMAN EMPIRE 27 point. The Senate required from him an oath that he would administer the Empire conscientiously and not visit offences upon anyone with whom he had had a quarrel. This exhibits the principle, which was con- stantly and chiefly applied for the purpose of preventing a new Emperor from making ecclesiastical innovations. It was a recognised condition of eligibility to the throne that the candidate should be a Christian, and an orthodox Christian. The latest pagan Emperor was Julian. After him it would have been virtually impossible for a pagan to rule at Constantinople. After the Council of Constantinople in a.d. 381, which crushed the Arian heresies, it would have been impossible for an Arian to wear the diadem. This was expressly recognised in the situation which ensued on the death of Theodosius II. The most prominent man at the moment was Aspar, but he was an Arian, and on that account alone his elevation was considered out of the question. Up to that period it may be said that such conditions of 28 THE CONSTITUTION OF faith were political rather than constitutional; but when the coronation ceremony was at- tended with religious forms, we may say that Christianity was coming to be considered a constitutional condition of eligibility. By religious forms, I do not mean the part which the Patriarch played in the act of coronation, which, as we have seen, had no ecclesiastical significance, but other parts of the ceremony, such as prayers, which were introduced in the fifth century. It was at the accession of Anastasius I that a religious declaration was first required from an Em- peror. Anastasius was with good reason suspected of heterodoxy ; he was in fact a monophysite. He was not asked to make any personal confession of faith, but at the Patriarch's demand, he signed a written oath that he would maintain the existing ecclesias- tical settlement unimpaired and introduce no novelty in the Church. We are ignorant whether such a written declaration was form- ally required at all subsequent elections; probably not ; but it was, we know, imposed THE LATER ROMAN EMPIRE 29 in a number of cases where there was reason to suspect a new Emperor of heretical ten- dencies. Ultimately, we cannot say at what time, this practice crystallised into the shape of a regular coronation oath, in which the monarch confesses and confirms the decrees of the Seven Ecumenical Councils and of the local synods, and the privileges of the Church, and vows to be a mild ruler and to abstain as far as possible from punishments of death and mutilation^^. The fact that such capitulations could be and were imposed at the time of election, even though the Emperor's obligation to submit to them was moral rather than legal, means that the autocracy was subject to limitations and was limited. But apart from such definite capitulations, the monarch's power was restricted by unwritten principles of government which bound him as much as the unwritten part of the English constitution binds our king and government. The autocrat was the supreme legislator; personally he was above the laws, solutus legibus^^; there 30 THE CONSTITUTION OF was no tribunal before which he could be summoned; but he was bound by the principles and the forms of the law which was the great glory of Roman civilisation 2^. He could modify laws, he could make new laws; but no Emperor ever questioned the obligation of conforming his acts to the law or presumed to assert that he could set it aside. Although theoretically above the law, he was at the same time bound by it, alligatus legihus, as Theodosius II expressly acknowledges^^. Basil I, in a legal handbook, explicitly afl&rms the obligation of the Emperor to maintain not only the Scriptures and the canons of the Seven Councils, but also the Roman laws. And the laws embraced the institutions. Though changing circum- stances led to adaptations and alterations, the Byzantine conservatism, which is almost proverbial and is often exaggerated, attests the strength of the unwritten limitations which always restrained the Imperial auto- cracy. The Senate, too, though it had no share THE LATER ROMAN EMPIRE 31 in the sovranty, might operate as a check on the sovran's actions. For there were various political matters which the Emperor was bound by custom to lay before it. We have not the material for enumerating what those matters were, but among the most im- portant were questions of peace and war and the conclusion of treaties. The Senate would obediently concur in the views of a strong sovran, and probably its meetings were generally of a purely formal nature, but it is signiiScant that in the case of a weak Emperor (Michael I) we find the Senate opposing the autocrat's wishes and the auto- crat bowing to their opinion^. It is implied in what I have said that the Church represented a limit on the Emperor's power. From the ninth century onward, the Decrees of the Seven Councils were an unalterable law which no Emperor could touch ^^. At the same time, the relation of the state to the Church, of which I must now speak, illustrates the amplitude of his power. The Byzantine Church is the most important 32 THE CONSTITUTION OF example in history of a State-Church. Its head was the Emperor. He was considered the delegate of God in a sphere which included the ecclesiastical as well as the secular order. The Patriarch of Con- stantinople was his minister of the depart- ment of religion, and though the usual forms of episcopal election were observed, was virtually appointed by him. It was the Emperor who convoked the Ecumenical Councils, and it was the Emperor who pre- sided at them either in person or, if he did not care to suffer the boredom of theo- logical debates, represented by some of his secular ministers^. Canonical decrees passed at councils did not become obligatory till they were confirmed by the Emperor; and the Emperors issued edicts and laws relating to purely ecclesiastical affairs, quite inde- pendently of Councils. The Patriarch Menas asserted in the reign of Justinian that nothing should be done in the Church contrary to the Emperor's will, and Justinian, who was the incarnation of sacerdotal monarchy, was THE LATER ROMAN EMPIRE 33 acclaimed as High-Priest Basileus (ayo^tepevs l3aG-i\evrj. The meaning of lex generalis (briefly, an edict promulgated as applicable to the whole Empire) is explained ih. 8, which is based on Cod. Just. i. 14. 3. The Emperor could not enact a special constitution, — applicable to a section, district, or town, — which was contrary to the provisions of a lex gene- ralis. ^ Cod. Just. i. 14. 4, digna vox maiestate regnantis legibus alligatum se principem profiteri : adeo de auctori- tate iuris nostra pendet auctoritas. 22 The functions of the Senate seem to have closely resembled those of the Synedrion in the Hellenistic king- doms. Compare the account of a meeting of the Synedrion of Antiochus in Polybius, v. 41-42. It may be noticed that during the minority after the death of Romanus II, it is the Senate that appoints Nicephorus II to the supreme command of the Asiatic troops (Leo Diaconus, ii. 12). The NOTES importance of the Senate is illustrated by the political measure of Constantino X who "democratized" it: see Psellos, Historia^ p. 238 (ed. Sathas, 1899) ; C. Neumann, Die Weltstellung des hyzcmtinischen Belches vor den Kreuzzugerij p. 79. 23 This principle had been already laid down by Justinian in regard to the first four Councils, the decrees of which he places on the same level as Holy Scripture : Nov. 151, a, ed. Zacharia, ii. p. 267. 24 The best general account of the relation of State and Church in Byzantium will be found in the late Professor Gelzer's article in the Historische Zeitschrift, N. F. vol. 50, 193 sqq. (1901). At the Seventh Ecumenical Council (a.d. 787) the presidency was committed to the Patriarch Tarasios, evidently because he had been a layman and minister, not (like most of his predecessors) a monk. 25 Augustinus Triumphus, Summa de potestate Ecch- siastica, i. 1, p. 2, ed. 1584 (Kome) : si quis eligatur in Papam nullum ordinem habens, erit verus Papa et habebit omnem potestatem iurisdictionis in spiritualibus et tem- poralibus et tamen nullam habebit potestatem ordinis. 26 Paraenesis ad Leonem^ in Migne, Pair, Gr, cvii. pp. XXV, xxxii. 27 See Photius, in Migne, P. G. cii. 765 and 573. Cp. Sickel, op. cit. 547-8, and on the other hand Bright- man, op. cit. 383-5. 28 The Holy Roman Empire (last ed. 1904), 343 sqq. 29 This is noted by Sidgwick, Development of European Polity, p. 10. ^ For an analysis of the conception of unconstitutional as distinguished from illegal see Austin, op. cit, 265 sqq. B. CTambriljge: PRINTED BY JOHN CLAY, M.A. AT THE UNIVEBSITY PBBS8.