Excerts from

Despite its extreme structural simplicity there was no ideal of squalor or poverty about the pre-Nicene celebration of the eucharist. The list of church plate at Cirta and many other such indications are a sufficient guarantee of that. The baptistery attached to the house-church at Dura-Europos (c. a.d. 230) was painted from floor to ceiling with pictures of scenes from the Old and New Testaments, and a similar decoration of the assembly-room of the church had just been begun when the building was destroyed. There could be a considerable degree of splendour about the setting of the ecclesia in a great Roman patrician house, and even where this was lacking attempts were evidently made to supply some dignity. There was no puritan cult of bareness for its own sake.

There was, too, an element of ceremony in the celebration and a good deal of moving about. The rite was viewed essentially as an action, and a number of people cannot combine to take different parts in a corporate action without some such element of ceremony, in the sense of organized and concerted movement. It was a large part of the deacon’s ‘liturgy’ by his ‘proclamations’ to direct and give the signal for these movements. There was, too, an element of solemnity; the bishop’s prayer was probably chanted as the Jewish prayers had been chanted. The use of the informal speaking voice for any part of the eucharist appears to be an innovation of the Latin churches in the early middle ages; for the eucharistic prayer itself it was not known before the Reformation. One cannot make much of the use by pre-Nicene writers of dicere (to say) in connection with the prayers. The ancients habitually used this word of a recitative, e.g. dicere carmen (lit. = ‘to say a song’). Probably the immemorial preface-chant of the West represents approximately the way in which the whole eucharistic prayer was originally recited there. Very similar intonations are traditional for the public prayers of the liturgy all over the East.

When all is said and done, the impression left by the early evidence about the celebration of the eucharist is one not so much of simplicity as of great directness, as became a deliberately ‘domestic’ act. There was no elaborate or choral music at the eucharist as at the synaxis; no special vestments or liturgical ornaments or symbolism, nothing whatever to arouse the emotions or stir the senses or impress the mind — just a complete and intense concentration upon the corporate performance of the eucharistic action in its naked self, without devotional elaborations of any kind whatever.

It is very easy for us to romanticize the life and worship of the primitive Christians. What was conventional in the social setting of their day has for us the picturesqueness of the strange and remote; what was straightforward directness in their worship has for us the majesty of antiquity. It is a useful thing occasionally to transpose it all into the conventions of our own day and look at the result.

Suppose you were a grocer in Brondesbury, a tradesman in a small way of business, as so many of the early Roman Christians were. Week by week at half-past four or five o’clock on Sunday morning (an ordinary working-day in pagan Rome) before most people were stirring, you would set out through the silent streets, with something in your pocket looking very like what we should call a bun or a scone. At the end of your walk you would slip in through the mews at the back of one of the big houses near Hyde Park, owned by a wealthy Christian woman. There in her big drawing-room, looking just as it did every day, you would find the ‘church’ assembling — socially a very mixed gathering indeed. A man would look at you keenly as you went in, the deacon ‘observing those who come in’ (Didascalidy ii. 57), but he knows you and smiles and says something. Inside you mostly know one another well, you exchange greetings and nod and smile; (people who are jointly risking at the least penal servitude for life by what they are doing generally make certain that they know their associates). At the other end of the drawing-room sitting in the best arm-chair is an elderly man, a gentleman by his clothes but nothing out of the ordinary — the bishop of London. On either side of him is standing another man, perhaps talking quietly to him. On chairs in a semicircle facing down the room, looking very obviously like what they are — a committee — sit the presbyters. In front of them is a small drawing-room table.

The eucharist is about to begin. The bishop stands and greets the church. At once there is silence and order, and the church replies. Then each man turns and grasps his neighbour strongly and warmly by both hands. (I am trying to represent the ancient by a modern convention. The kiss was anciently a much commoner salutation than it is with us in England, but it implied more affection than does merely ‘shaking hands’ with us.) The two men by the bishop spread a white table-cloth on the table, and then stand in front of it, one holding a silver salver and the other a two-handled silver loving-cup. One by one you all file up and put your little scones on the salver and pour a little wine into the loving-cup. Then some of the scones are piled together before the bishop on the cloth, and he adds another for himself, while water is poured into the wine in the cup and it is set before him. In silence he and the presbyters stand with their hands outstretched over the offerings, and then follow the dialogue and the chanted prayer lasting perhaps five minutes or rather less. You all answer ‘Amen’ and there follows a pause as the bishop breaks one of the scones and eats a piece. He stands a moment in prayer and then takes three sips from the cup, while the two men beside him break the other scones into pieces. To each of those around him he gives a small piece and three sips from the cup. Then with the broken bread piled on the salver he comes forward and stands before the table with one of the deacons in a lounge suit standing beside him with the cup. One by one you file up again to receive in your hands ‘The Bread of Heaven in Christ Jesus,’ and pass on to take three sips from the cup held by the deacon, ‘In God the Father Almighty and in the Lord Jesus Christ and in the Holy Spirit in the holy church,’ to which you answer ‘Amen’; then you all file back again to where you were standing before. There is a moment’s pause when all have finished, and then most of you go up to the bishop again with a little silver box like a snuff-box into which he places some fragments of the Bread. You stow it in an inside pocket, reflecting perhaps that Tarcisius was lynched six months ago for being caught with one of those little boxes upon him. There is another pause while the vessels are cleansed, and then someone says loudly ‘That’s all. Good morning, everybody.’ And in twos and threes you slip out again through the back door or the area door and go home — twenty minutes after you came in. That is all there is to it, externally. It would be absolutely meaningless to an outsider, and quite unimpressive.

But perhaps it did not all end quite so easily. You might very well never walk back up Maida Vale again. Perhaps the bishop stopped to speak to someone on the front-door steps as he went out, and was recognised by a casual passer-by who set up a great shout of ‘Christian! Christian!’ And before anyone quite realised what was happening a small jostling crowd had collected from nowhere and someone had thrown a brick through one of the windows; doors and windows were opening all down the street and there was a hubbub of jeers and yells, till a policeman arrived majestically, demanding ‘Wot’s all this ‘ere?’ ‘It’s those --- Christians again!’ shouts someone, and the policeman gets out his notebook and looks severely at the bishop standing with the two deacons just behind him at the foot of the steps. ‘Wot’s all this about?’ And then in response to the accusing shouts of the elbowing crowd there comes the deadly challenge from the policeman, ‘Is that right that you’re a Christian?’ And the bishop admits he is a Christian. ‘There’s another of them,’ says someone, pointing at one of the deacons. ‘There’s a whole gang of them in there.’ The deacons briefly admit their faith, and the policeman looks doubtfully at the house. It’s said that they always come quietly, but one never knows. He blows his whistle, more police arrive, the house is entered, and soon afterwards twenty-two people, including the bishop and his deacons and the little grocer from Brondesbury, are marched off to the station.

The proceedings are by summary jurisdiction, as in the case of a raid on a night-club with us. They are all charged together ‘with being Christians,’ i.e. members of an unlawful association. Each is asked in turn whether he pleads guilty or not guilty. If he answers ‘guilty,’ his case is virtually decided. The magistrate is perfectly well aware of the Christian rule of never denying their religion. Someone’s courage fails at the critical moment and he falters ‘Not guilty.’ Then there is a simple further test to be applied. At the side of the court-room is hung a picture of the king. Just go and kneel in front of that picture and say "Lord have mercy upon me," will you?’ says the magistrate. (The offering of the conventional pinch of incense or few drops of wine before the statue of the deified emperor, which was the routine test for Christianity, involved no more religious conviction than such a ceremony as I have invented here.) Some of the accused go through the prescribed test with white faces and faltering lips. One goes to the picture to do so and his conscience suddenly gets the better of his fear; he knocks the picture off the wall in a revulsion of nervous anger. He is hustled back to the dock and the picture is hung up again. The magistrate, a reasonable man, again asks each of those who have pleaded guilty whether they will even now go through the little ceremony. They all refuse. There is no more to be done, no possible doubt as to the law on the matter: non licet esse christianos; Christians may not exist.’ The legal penalty is death, and there is no ground of appeal. As a rule there is no delay. Unless they were reserved for the arena, sentences on Christians were usually carried out on the same day. So in our modern analogy fifteen Christians were hanged that afternoon at Wandsworth. On other occasions the policy of the administration might have caused private instructions to be issued to the magistrates that the law against Christianity is not to be too strictly enforced for the present; a sentence of the ‘cat’, penal servitude for life and transportation would have been substituted for the death-penalty. Whether this was really much more merciful may be doubted. The imperial lead-mines in Sardinia, for instance, which were the usual convict-station for Roman Christians in such a case, must have been even more like Devil’s Island than Botany Bay. Most of the prisoners died within two or three years.

We shall not begin to understand what the eucharist meant to Christians until we have estimated this background of real danger and intense hatred in a setting of absolutely normal daily life. It is true that organised and official persecution by the state was by no means continuous, that there were long periods when the central government was otherwise occupied, and wide regions where the local authorities were inclined to turn a blind eye to the existence of Christians, provided these did not thrust themselves upon their notice. But there were other periods and equally wide regions where official persecution raged with violence for years together. For two hundred years, from Nero to Valerian (roughly a.d. 65-260), Christian worship was in itself a capital crime. For another fifty after that, the law against Christian assembly relaxed; but to be a Christian was, by an illogicality, still brought under the capital charge of laesa maiestas. There is the opinion of Ulpian the jurist and the actual contemporary court-record of martyrdoms to prove that even in this period of peace in the latter half of the third century martyrdom was still only a matter of whether you happened to be accused. No one ever knew even in a period when the government was quiescent when persecution might not break out in the form of mob-violence, or what trivial cause might bring upon a man the inescapable official challenge ‘Art thou a Christian?’ Callistus trying to recover a commercial debt from Jewish debtors finds them making this charge against him in the prefect’s court to avoid payment; and within an hour or two he has been scourged and sentenced for life to the deadly Sardinian mines (Hippolytus, Philosophumendi ix. II). Marinus, the soldier accused of Christianity by a comrade envious of his promotion to centurion, is dead three hours after the accusation has been lodged (Eusebius, Eccl. Hist., VII. xv. 1). Both these typical stories are reported by contemporaries from periods which rank more or less as times of toleration. We can and should distinguish between the intermittent hostility of the government and the unorganized and unpredictable malignity of the mob or of private informers. But when all has been said that is true in mitigation of the severity of ancient persecutions, for two hundred and fifty years from Nero to Constantine to be a Christian was in itself a capital crime, always liable to the severest penalty, even when the law was not enforced. It remains a demonstrable historical fact from contemporary records that during this period thousands of men and women were killed, tens of thousands more suffered grievously in their fortunes and persons, and hundreds of thousands had to put up with the opposition of their families and the suspicion and ostracism of their neighbours for half-a-lifetime and more. And the storm center throughout the whole period was undoubtedly the eucharist.

When we regard what actually took place in the early eucharistic rite, the fear and hatred it inspired over so long a time seem ridiculous. Yet it is an uncanny fact that there is still scarcely any subject on which the imagination of those outside the faith is more apt to surrender to the unrestrained nonsense of panic than that of what happens at the catholic eucharist. As a trivial instance, I remember that my own grandmother, a devout Wesleyan, believed to her dying day that at the Roman Catholic mass the priest let a crab loose upon the altar, which it was his mysterious duty to prevent from crawling sideways into the view of the congregation. (Hence the gestures of the celebrant.) How she became possessed of this notion, or what she supposed eventually happened to the crustacean I never discovered. But she affirmed with the utmost sincerity that she had once with her own eyes actually watched this horrible rite in progress; and there could be no doubt of the deplorable effect that solitary visit to a Roman Catholic church had had on her estimate of Roman Catholics in general, though she was the soul of charity in all things else. To all suggestions that the mass might be intended as some sort of holy communion service she replied only with the wise and gentle pity of the fully-informed for the ignorant.

I mention this peculiar opinion of a good and sensible woman because it illustrates well enough a frame of mind among the ancient pagans which was at once a cause and a result of Christian secrecy about the eucharist. The gruesome stories of ritual murder and cannibal feasts which have been told since the stone age — when, no doubt, they had their justification — about all unpopular associations, received a fresh impulse from misunderstandings of indiscreet Christian talk of receiving ‘the Body and the Blood.’ The dark suspicions of orgies of promiscuous vice or even organised incest, which the nasty side of men’s imaginations is always willing to credit about mysterious private gatherings, were stimulated by talk of ‘the kiss’ and of ‘brothers’ and ‘sisters.’ The point is that these charges against the Christians were taken with the utmost seriousness by multitudes not only of the cruel and foolish and ignorant but of normally humane and sensible men. When the heathen slaves of a Christian master broke down under the torture always employed in the Roman courts to ensure the truthfulness of a slave’s evidence — such was the extraordinary reason seriously maintained for the practice — and proceeded to ‘confess’ their knowledge of such goings on among the Christians, it may have added to the disgust with which the decent pagan regarded all mention of the eucharist, but hardly at all to the strength of the general conviction that the holding of the ecclesia ought to be stopped by the authorities at all costs. One has only to read, for instance, the account by an eye-witness at Lyons in a.d. 177 of the pathetic occasion in the persecution there when after just such a ‘confession’ by heathen slaves the apostate Christians were mobbed by the crowd as self-confessed ‘polluted wretches’ (miarous), to realise just what associations the very word ‘eucharist’ would have in the mind of any decent Lyonnais for the next thirty years, or what sort of hysteria a rumour of the holding of Christian worship would be likely to work up in the city.

The imperial government was a great deal better informed than the populace. It regarded the church as a potential political danger for precisely the same reasons as any other totalitarian government is bound to do so. At times it took vigorous measures to protect itself against this danger, and it is an instance of Roman governmental capacity that whenever it did so it showed a clear understanding of the problem which confronted it. Active measures were always directed not so much against the holding of Christian beliefs as against the expression of that belief in the worship of the ecclesia. Those officials, for instance, who actually carried out the persecution under the emperor Decius (a.d. 250-251) must have been perfectly well aware from their behaviour that of the thousands of Christian apostates who offered sacrifice under threat of instant martyrdom, the vast majority remained sincerely convinced Christians in belief, even though by the failure of their courage at the moment of trial they now faced life-long exclusion from Christian communion. The persecutors were not concerned to produce sincere believers in the deity either of the emperor or of the Olympian gods, but to put an end to the illegal meetings of the Christian ecclesia. They could be content with the merest pretence of conformity because they could rely on the discipline of the church itself to exclude from the ecclesia all who had in any way compromised. The government’s attack was pressed all the time upon worship, by striking especially at the clergy with martyrdom or penal servitude, by the confiscation of all property upon which Christian worship was proved to have taken place, and by a variety of other measures, all designed to make impossible the holding of the ecclesia. But there was no parallel attempt by a counter-propaganda to discredit Christian beliefs or to defend pagan ones.

The church being what it was, the act of taking part in the common worship could be accepted by church and state alike as the effective test of Christianity. From the point of view of the state it was deliberate treason (laesa maiestas). From the point of view of the church the corporate action of the eucharist in the ecclesia was the supreme positive affirmation before God of the Christian life. There was no place on either view for that modern ‘Christianity’ which owns no allegiance to the church and her worship. To the state an academic belief which did not express itself in worship carried no danger of Christian allegiance. To the church belief which did not express itself in worship would have seemed both pointless and fruitless. Christian belief was the condition of admission to that worship, explicitly required before baptism and confirmation, which alone admitted a man to pray with the church, let alone communicate. On the other hand, for a confirmed Christian to allow himself to take any part whatever in non-christian worship was ‘apostasy,’ a public declaration that he renounced that faith in Christ as his redeemer which was his passport to worship. Down to a.d. 252 apostasy involved perpetual exclusion from the ecclesia in this world and damnation in the next, unless perhaps the lapsed Christian might hope to move the mercy of God after death by a life-long penance outside the corporate life of the church. The state was content to accept the logic of the Christian principle that religious belief can only be finally and adequately expressed by worship. When the well-organised Decian persecution encouraged apostasy by making compliance easy, and reaped an immense harvest of lapses, it must have seemed that the church was about to be strangled in her own inviolable discipline.

The church met the crisis by a revolutionary change in that discipline, which the government does not seem to have anticipated. In the teeth of bitter opposition from the zealots everywhere, the bishops restored to membership of the ecclesia all apostates who showed the sincerity of their repentance by undergoing a period of penance. The lapsed flocked back in thousands, and the correspondence of S. Cyprian contains abundant evidence with what eagerness they sought to resume their Christian life, not as believers — they had never ceased to be that — but as worshippers. For the Christian as for the persecutor the liturgy formed the very life not only of the church corporately but of the individual soul. It was a statesmanlike move, probably the only one which could have enabled the church to survive the second wave of persecution which the baffled government at once launched against the Christian revival under Valerian (a.d. 254-9). The state was eventually distracted by foreign war, and had to own itself unable to stamp out the ecclesia. An edict of Gallienus conceded permission to the Christians freely ‘to use their ecclesiai’ the property in which was restored to them (a.d. 260).

This was a virtual concession of freedom of worship, but it left the legal position ambiguous. Christian worship was no longer in itself a crime, and the church became a tolerated if not a legally recognised association. But Christianity was not a legal religion, and the individual Christian could still be charged with high treason.

For the next forty years the state simply turned its back upon the fact that the church existed, though everyone was aware that ‘the Christian question’ would have to be faced one day. But the forty years of uneasy toleration which ended the third century brought a considerable increase in Christian numbers, which together with the liberty of assembly now permitted, began to force upon the church a more regular organization of her worship. We find special church buildings for this purpose beginning to be erected in many towns and even in some quarters of Rome itself during this period. In Asia Minor especially the church came to number quite a large proportion of the population and could come more into the open. At Nicodmeia, the Eastern capital, where high officers of the court and even members of the royal family were attracted to the church, the Christian bishop’s cathedral is said to have been the most imposing public building in the city before the end of the third century.

Elsewhere, Christians were usually an unpopular minority, and worship had to be conducted with more discretion. But everywhere (as we have seen at Cirta) it was now an open secret where Christian worship was held and who the Christian clergy were. When the last tempest of persecution arose under Diocletian a.d. 303-13 — the longest as well as the fiercest the church ever had to face — it was again upon Christian worship that it pressed most fiercely. That worship was itself now much more open to attack by reason of its new semi-public organization. This time, too, there was a real attempt to refute Christian teaching by intellectual propaganda, and a systematic destruction of Christian literature. The virtual prevention of corporate worship except in the most furtive fashion for nearly ten years and the gradual extinction of the clergy by martyrdom or apostasy did on this occasion reduce the church to the direst extremities, in a way no previous persecution had ever done. The edicts of toleration put out in 313 by the emperors Maximin and Maximian, and comprehensively ratified and enforced by the new Christian emperor Constantine in the following year, came only just in time to save her from complete disorganization. The West was now finally free from organised persecution by the state, but the Eastern provinces still had to endure it intermittently for another five years

It will be seen that popular and official persecution of the church had very different motives. The state feared the church; the populace disliked the Christians. The state wished to make apostates; the mob as a rule preferred martyrs. It is a constant feature of the genuine Acta of the martyrs to find the magistrate arguing and pleading with the prisoner to deny his faith and fulfil the formal test of sacrifice, even delaying and straining the law sometimes to secure something which will pass for a denial, while the mob howl for the prisoner’s death.

The Roman judicial standard was on the whole a high one. There is evidence that many of the magistrates did not enjoy the duty of enforcing the law against Christians, and recognised its futility and injustice. But though the administration might often be disposed to avoid charging men with Christianity, the law placed a fatal weapon in the hands of both the hostility of the mob and private enmity. Once the accusation of Christianity had been brought to his notice the magistrate was bound to take cognizance of it. And once a man was put that fatal question ‘Art thou a Christian?’ there was no other way but apostasy or sentence. The magistrate and the martyr were alike helpless. It was always open to a magistrate more energetic or fanatical than his fellows to set the law in motion himself within his jurisdiction. But except when instructions were received from the central administration to ‘tighten things up,’ this appears to have been comparatively rare; and the general practice of changing the local magistrates annually usually ensured a brief duration to such local official action.

It is plain from second and third century Christian literature that the great permanent danger to the Christians came from the mob. As Tertullian puts it, ‘They think the Christians are at the bottom of every disaster to the state and every misfortune of the people. If the Tiber floods the city or the Nile fails to flood the fields, if there are portents in heaven or earthquakes on earth, if famine comes or plague, they clamour instantly "Throw the Christians to the lion." So many, to one lion?’ (Apologeticus, xi).

Thus the church could not meet the charges of cannibalism and incest, which the man in the street honestly believed about the eucharist, in the only way which might have been effective — though it did not convince my grandmother — by holding the rite with absolute publicity. This was partly at least because the state made the holding of Christian worship in itself a 1 Tertullian, capital crime. In any case she would probably have been reluctant to do this in a pagan worlds because the eucharist expressed in its very essence and idea the ‘separateness’ of the holy church from ‘the world that lieth in wickedness’ (John 5:19).

There was thus left only the alternative of denying the charges as often as possible in the course of propaganda, and enduring their consequences when this failed — as it invariably did — to convince the public. Justin in the famous ‘Open Letter to the Government’ which is known as his First Apology tried the expedient of describing just what was done at the eucharist with a disarming frankness, which to a modern reader must seem a convincing (and rather skilful) demonstration of its entire harmlessness. Yet it had no effect whatever on contemporary opinion. In his second manifesto of the same kind issued a year or two later, Justin himself obviously despairs of achieving much by this method of reasonableness, and adopts a much more indignant and defiant tone.

Tertullian used instead the method of a biting irony. But it is obvious throughout the book that though he addresses the administration he is really trying to counter the popular rumours about orgies at the eucharist, which are having a very serious effect. He twits the officials with the fact that they have never been able to discover the scantiest factual evidence for these charges — ‘how many babies any particular person has eaten, how many times he has committed incest, who the cooks were.... What a boast for any governor, if he had actually caught a man who had eaten a hundred babies!’ (Tertullian, Ap. ii). But his argument on these things is really addressed not to the officials, who did not take these charges seriously, but to the public which did. ‘Suppose these things are true for the moment. I only ask you who believe that such things are done to imagine yourself eager for the eternal life they are supposed to secure. Now! Plunge your knife into an innocent baby that never did anyone any harm, a foundling. Perhaps that is some other Christian’s office. Well, any way, stand looking down on this human being gasping in death almost before it has lived; wait while its new little soul escapes; catch its gurgling blood and soak your bread in that. Then gulp it down with pleasure! Then lie down and point out where your mother is to lie and where your sister. Take careful note, that when the dogs (chained to the lampstand) plunge all in darkness you may make no mistake. You will have done a sacrilege if you fail to commit incest. By these mysteries and this confirmation you shall live for all eternity. Tell me, now, is eternity worth that?.. Even if you thought so, I deny that you would want it on those terms. Even if you did want it, I deny that you could bring yourself to gain it thus. Why then can others, if you cannot? Why can you not, if others can? We are different from you in nature, I suppose — dog-headed men or sciapods? We have a different sort of teeth, or feel a different lust? You believe men can do these things? Then presumably you can do them. You are a man yourself, just like a Christian. If you know you could not bring yourself to do them, then do not believe that others can.... I suppose when someone wants to be initiated in this way he first goes to the high-priest of these mysteries, to find out what preparations he must make. And he tells him, "Oh, you will need a baby, a teeny baby, which does not understand death and will smile under your knife; and bread in which to catch its squirting blood ... and above all, you must bring your mother and your sister." What if they will not come, or the convert has none? What about Christians who have no near feminine relations? I presume he can be no rightful Christian unless he be a brother or a son?’ (Tertullian, Ap., viii).

Of course this sort of firework did no more good than Justin’s calculated naiveté. Indeed Tertullian’s whole Apology is so much in the nature of a devastating counter-attack on paganism all along the line that it seems more calculated to infuriate any conventionally-minded pagan who happened to read it than to soothe his alarm at the alleged revolutionary opinions and morals of the Christians. But the lurid background of suspicion and calumny about the eucharist and ill-will towards those who took part in it has to be borne in mind in considering the importance that Christians attached to its celebration and the reasons why they clung to this ill-famed rite.

These men and women did not run continual risks to attend it merely because there they remembered with thankfulness in a specially moving way the death of Jesus which had redeemed them. They could do that anywhere and alone; some of them did it most of their waking hours. Nor was it simply that in the eucharist alone they could satisfy a personal longing for God by receiving holy communion. As a matter of fact if a devout third century Christian on his deathbed could have reckoned up all the communions he had ever made, he would probably have found that the large majority had been made from the reserved sacrament at home, quite apart from the liturgy. These desires of Christian personal devotion could be and were satisfied in private in comparative safety, without the dangers and scandal which centered round the eucharist. There was, indeed, a rather striking absence from the primitive eucharistic rite of any devotional practice which was calculated to arouse or feed a subjective piety — no confession of sins or devotions in preparation for communion, no corporate thanksgiving even, nothing but the bare requisites for the sacramental act. It was a burning faith in the vital importance of that eucharist action as such, its importance to God and to the church and to a man’s own soul, for this world and for the next, which made the Christians cling to the rite of the eucharist against all odds. Nothing else could have maintained the corporate celebration of the liturgy through the centuries when the ecclesia was outside the law.

For these Christian men and women were very normal. They were not impossibly heroic. Their answers in the dock often shew that they were very frightened. Even when they were most defiant their rudeness is often a mark of fear. Few men could look forward to the appalling tortures which the courts in the later second century sometimes took to applying — ‘to make them deny their crime’ as Tertullian bitterly remarked, ‘not like other criminals to confess it’ — without considerable perturbation. Many of them apostatised when it came to the final test, often most of them. The world, the flesh and the devil were as active and deadly with them as they are with Christians nowadays. And so was another enemy whose assaults on the church of the martyrs we often ignore though we know its deadening effects on ourselves — routine, the mere fact that one has been trying to be a Christian for quite a long time and little seems to come of it. The parable of the Sower was just as true then as now. But these normal men and women were prepared with open eyes to accept the risks and inconveniences they undoubtedly did encounter, just to be present at the eucharist together and regularly. I submit that it casts a flood of light on their beliefs about the eucharist and the nature of the church and Christian salvation generally, that they attributed this desperate importance not so much to ‘making their communion’ as to taking part in the corporate action of the eucharist.

It was to secure the fulness of this corporate action that a presbyter and a deacon had to be smuggled somehow into the imperial prisons, there to celebrate their last eucharist for the confessors awaiting execution; and S. Cyprian takes it as a matter of course that this must be arranged (Cyprian, Ep., v. 2). To secure this for his companions as best he could, the presbyter Lucian lying with his legs wrenched wide apart in the stocks of the prison at Antioch celebrated the mysteries for the last time with the elements resting on his own breast, and passed their last communion to the others lying equally helpless in the dark around him. To secure this a whole congregation of obscure provincials at Abilinitina in Africa took the risk of almost certain detection by assembling at the height of the Diocletian persecution in their own town, where the authorities were on the watch for them, because, as they said in court, the eucharist had been lacking a long while through the apostasy of their bishop Fundanus, and they could no longer bear the lack of it. And so they called on a presbyter to celebrate — and paid the penalty of their faith to a man. To secure this was always the first thought of Christians in time of threatened persecution. ‘But how shall we meet, you ask, how shall we celebrate the Lord’s solemnities? ... If you cannot meet by day, there is always the night,’ says Tertullian, bracing the fearful to stay and meet the coming storm. Even when a church had been scattered by long persecution, the duty was never forgotten. ‘At first they drove us out and ... we kept our festival even then, pursued and put to death by all, and every single spot where we were afflicted became to us a place of assembly for the feast — field, desert, ship, inn, prison,’ writes S. Denys, bishop of Alexandria, of one terrible Easter day c. a.d. 250, when a raging civil war, famine and pestilence were added to the woes of his persecuted church.

Literally scores of similar illustrations from contemporary documents of unimpeachable historical authority are available of the fact that it was not so much the personal reception of holy communion as the corporate eucharistic action as a whole (which included communion) which was then regarded as the very essence of the life of the church, and through that of the individual Christian soul. In this corporate action alone each Christian could fulfil for himself or herself the ‘appointed liturgy’ of his order, and so fulfil his redeemed being as a member of Christ. For my own part I have long found it difficult to understand exactly how the eucharist ever came to be supposed by serious scholars at all closely comparable with the rites of the pagan mysteries. The approach is so different. In the mysteries there is always the attempt to arouse and play upon religious emotion, by long preparation and fasts, and (often) by elaborate ceremonies, or by alternations of light and darkness, by mystical symbols and impressive surroundings, and pageantry; or sometimes by the weird and repulsive or horrible. But always there is the attempt to impress, to arouse emotion of some kind, and so to put the initiate into a receptive frame of mind. As Aristotle said, men came to these rites ‘not to learn something but to experience something.’ The Christian eucharist in practice was the reverse of all this. All was homely and unemotional to a degree. The Christian came to the eucharist, not indeed ‘to learn something,’ for faith was presupposed, but certainly not to seek a psychological thrill. He came simply to do something, which he conceived he had an overwhelming personal duty to do, come what might. What brought him to the eucharist week by week, despite all dangers and inconveniences, was no thrill provoked by the service itself, which was bare and unimpressive to the point of dullness, and would soon lose any attraction of novelty. Nor yet was it a longing for personal communion with God, which he could and did fulfil otherwise in his daily communion from the reserved sacrament at home. What brought him was an intense belief that in the eucharistic action of the Body of Christ, as in no other way, he himself took a part in that act of sacrificial obedience to the will of God which was consummated on Calvary and which had redeemed the world, including himself. What brought him was the conviction that there rested on each of the redeemed an absolute necessity so to take his own part in the self-offering of Christ, a necessity more binding even than the instinct of self-preservation. Simply as members of Christ’s Body, the church, all Christians must do this, and they can do it in no other way than that which was the last command of Jesus to His own. That rule of the absolute obligation upon each of the faithful of presence at Sunday mass under pain of mortal sin, which seems so mechanical and formalist to the protestant, is something which was burned into the corporate mind of historic Christendom in the centuries between Nero and Diocletian. But it rests upon something more evangelical and more profound than historical memories. It expresses as nothing else can the whole New Testament doctrine of redemption; of Jesus, God and Man, as the only Saviour of mankind, Who intends to draw all men unto Him by His sacrificial and atoning death; and of the church as the communion of redeemed sinners, the Body of Christ, corporately invested with His own mission of salvation to the world.

Despite all the formalism and carelessness and hypocrisy which a social tradition of the general attendance at the eucharist of all who have been baptised involves, and has always involved, in catholic countries, there is this to be said: that no personal subjective devotion on the part of select individual communicants can manifest Christ as the redeemer of all men and of all human life, either to themselves or to the world or before God. Nor can the corporate being of the church as His one Body with many members be fulfilled in an action from which the greater part of the baptized and confirmed members are regarded or regard themselves as tacitly excluded.

We do well to approach the mystery of Christ’s Body and Blood with the profoundest reverence and searching of heart. Yet a eucharist where the table is ‘fenced,’ even only by the consensus of Christian opinion, a eucharist at which frequency has come to be regarded as a special preserve of the clergy and ‘the devout,’ and at which the majority of practicing Christians are present only on comparatively rare occasions — this has just as much ceased to be the scriptural and primitive eucharist as has the most unprayerful and conventional non-communicating attendance at Sunday mass by the tradesmen of a Sicilian country town.

The unfamiliarity of a vast proportion of ‘C. of E.’ Christians with the eucharist may have begun with a false notion of reverence. It has ended by destroying the true understanding of the eucharist even among many of those who still frequent it. The clergy will all have encountered those choice souls who actually prefer to ‘make their communion’ only in the peace of a week-day celebration, where three or four leisured people can scatter themselves widely all over the church, and avoid disturbance by the larger congregation at ‘the 8 o’clock’ on Sunday. It would probably surprise the clergy to find how widespread this self-centred devotion is among the laity, and how many regular communicants would prefer to fulfil their personal religious needs in this way if their situation gave them the weekday leisure. This is not much better than a parody of devotion to the eucharist, which our practice and teaching have somehow succeeded in implanting as the ideal. Behind it lie centuries of the mediaeval distortion of the eucharist as the focus of a subjective individual piety. In reality it is the very action of Him who came ‘to die not for that nation only, but that also He should gather together in one the children of God who were scattered abroad (John 11:52).

Let us look back for a moment. We have seen that the eucharist is primarily an action, our obedience to our Lord’s command to ‘Do this’; and that this action is performed by the Shape of the Liturgy, the outline of the service viewed as a single continuous whole. We have also seen that the meaning of this action is stated chiefly in the great eucharistic prayer, which formed the second item of that Tour-action shape’ of the eucharist which has come down almost from apostolic times. Since this prayer was originally ‘the’ prayer, the only prayer in the whole rite, it was there that the whole meaning of the rite had to be stated, if it was to be put into words at all in the course of the service. We have also noted that, while the tradition as to the outline of the rite was always and everywhere the same, there was no such original fixity about the content and sequence of this prayer. Its text was subject to constant development and revision, so that it varied considerably from church to church and from period to period, and even (probably within narrower limits) from celebrant to celebrant.

In this chapter we shall set out the oldest specimens of ancient local traditions of this prayer which have come down to us, together with other material which throws light upon them.

The traditions we shall chiefly consider now are three — those of Rome, Egypt and Syria, for Rome, Alexandria and Antioch were the three most important churches in pre-Nicene times. But there were other traditions of the prayer elsewhere, some of them equally ancient, in North Africa, Spain and Gaul in the West, and in the apostolic churches of the Balkans and Asia Minor in the East. Unfortunately, by the accidents of history it happens that no texts of the eucharistic prayers of these churches have survived from pre-Nicene times, or indeed from any period at which their evidence can usefully serve for even a tentative comparison with the really ancient material. Our survey is thus bound to be very incompletely representative of the whole liturgical wealth of the pre-Nicene church as it actually existed, and the reader may reasonably wonder how it would be affected if these lost traditions could be included. I believe that the answer is ‘very little in principle and a great deal in detail,’ because of the form of the conclusions to which the extant material actually leads. The missing traditions of the prayer, if they could be recovered, would probably shew in its structure and phrasing a diversity equal to, or even greater than, those which survive. Such little evidence as we have about them suggests that they were verbally as independent of the prayers which we do know as these clearly are of one another. On the other hand this fragmentary evidence, and still more the incidental statements about the eucharist in the writers from these churches, suggest equally strongly that their fundamental understanding of the rite, that ‘meaning’ of it which their eucharistic prayers sought to state, was the same in all essentials as that found in the prayers which have survived. Diversity of form and a fundamental identity of meaning seem to have been the marks of the old local tradition everywhere.

We begin once more with the Apostolic Tradition of Hippolytus, the most important source of information we possess on the liturgy of the pre-Nicene church. This invaluable document contains the only pre-Nicene text of a eucharistic prayer which has reached us without undergoing extensive later revision. We have to be on our guard, however, against interpreting all the other evidence exclusively in the light of this single document (which raises almost as many fresh problems as it solves, from one point of view), just because it is in this way of such unique interest and importance. In itself it represents only the local tradition of Rome, though at an early stage, before developments had become complicated.

After the opening dialogue, already sufficiently commented, Hippolytus’ prayer runs thus:

1. We render thanks unto Thee, O God, through Thy Beloved Servant Jesus Christ, Whom in the last times Thou didst send (to be) a Saviour and Redeemer and the Angel of Thy counsel; Who is Thy Word inseparable (front Thee);
2. through Whom Thou madest all things and in Whom Thou wast well-pleased;
3. Whom Thou didst send from heaven into the Virgin’s womb, and Who conceived within her was made flesh, and demonstrated to be Thy Son, being born of Holy Spirit and a Virgin;
4. Who fulfilling Thy will and procuring for Thee an holy people, stretched forth His hands for suffering (or for the passion) that He might release from sufferings them who have believed in Thee;
5. Who when He was betrayed to voluntary suffering (or the passion) in order that He might abolish death and rend the bonds of the devil and tread down hell and enlighten the righteous and establish the ordinance and demonstrate the resurrection,
6. taking bread (and) making eucharist to Thee, said: Take, eat; this is My Body, which is [or will be] broken for you.

(f) Likewise also the cup, saying: This is My Blood which is shed for you.

7. When ye do this ye do [or make ye] My ‘anamnesis.’
8. Now, therefore, doing the ‘anamnesis’ of His death and resurrection
9. we offer to Thee the bread and cup
10. making eucharist to Thee because Thou hast made us worthy to stand before Thee and minister as priests to Thee.
11. And we pray Thee that [Thou wouldest send Thy Holy Spirit upon the oblation of Thy holy church]1 Thou wouldest grant to all who partake to be made one, that they may be fulfilled with (the) Holy Spirit for the confirmation of (their) faith in truth;
12. that we may praise and glorify Thee through Thy Servant Jesus Christ through Whom honour and glory (be) unto Thee with (the) Holy Spirit in Thy holy church, now and for ever and world without end.

R/ Amen.

We may analyze the structure of the prayer thus:

1. Address: Relation of the Father to the Eternal Word.
2. Thanksgiving for Creation through the Word.
3. Thanksgiving for the Incarnation of the Word.
4. Thanksgiving for Redemption through the Passion of the Word.
5. Statement of Christ’s purpose in instituting the eucharist.
6. Statement of His Institution of the eucharist.
7. Statement of His virtual command to repeat the action of (g) with a virtual promise of the result attaching to such repetition.
8. Claim to the fulfillment of the promise in (g).
9. Offering of the elements
10. constituting obedience to the command in (g), with an interpretation of the meaning understood by this obedience.
11. Prayer for the effects of communion.
12. Doxology.

This prayer was written down more or less verbally in this form at Rome c. a.d. 215, but the author emphatically claims that it represents traditional Roman practice in his own youth a generation before. It appears certainthat some of the phrasing in a-e is of his own composition, and represents his own peculiar theology of the Trinity; and it is at least possible that the wording of other parts of the prayer is from his own pen. But this does not make it improbable that the structure of the prayer as a whole (including a-e) and some of its actual wording were really traditional at Rome. The following parallels from the writings of Justin Martyr (Rome c. a.d. 155) all occur in professedly eucharistic passages, and some are even more remarkable in Greek than in English for the resemblance of then-phrasing to that of Hippolytus.

(a) The bishop ‘sends up praise and glory to the Father of all through the Name of the Son and the Holy Ghost’ (Ap. 1:65).

(Jesus is the ‘Beloved,’ the ‘Servant,’ the ‘Saviour,’ the ‘Redeemer’ and the ‘Angel of God’s counsel’ in a number of passages in Justin, though none of them are explicitly about the eucharistic prayer; the Word is ‘not separable’ from the Father (Dialogue, 128) but again this is not explicitly connected with the eucharistic prayer.)

(b-d) The eucharist was instituted ‘that we might at the same time give thanks to God for the creation of the world with all that is therein for man’s sake, and for that He has delivered us from the evil wherein we were born, and for that He loosed (the bonds) of powers and principalities with a complete loosing by becoming subject to suffering according to His own wil’ (Dialogue, 41).

(c, d, g) ‘As by the Word of God Jesus Christ our Saviour was made flesh and had flesh and blood for our salvation, so also we have been taught that this food "eucharistised" by a formula of prayer which comes from Him ... is the flesh and blood of that Jesus Who was made flesh. For the apostles in the memoirs which are by them, which are called "gospels," have recorded that thus it was commanded them (to do): that Jesus took bread and gave thanks and said "Do this for the anamnesis of Me: This is My Body"; and likewise took the cup and gave thanks and said "This is My Blood"‘ (Ap. 1:66).

(h) ‘The offering of fine flour ordered (in the Old Testament) to be offered on behalf of those who were cleansed from leprosy was a type of the bread of the eucharist, which Jesus Christ our Lord ordered to be done [or ‘sacrificed’] for an anamnesis of His passion which He suffered on behalf of men, whose souls have (thereby) been cleansed from all iniquity’ (Dialogue, 41).

(i) ‘The sacrifices which are offered to God by us gentiles everywhere, that is the bread of the eucharist, and the cup likewise of the eucharist’ (Dialogue, 41).

(j) The bishop ‘sends up eucharists (thanksgivings) that we have been made worthy of these things by Him’ (Ap. I:65). ‘We (christians) are the true high-priestly race of God ... for God accepts sacrifices from no one but by the hands of His own priests’ (Dialogue, 116).

(k, l) These have no verbal parallels in Justin’s allusions to the eucharist like the above, though the same sentiments are to be found at large in his works.

We can thus at the least say that there is nothing whatever in the specifically eucharistic teaching of Hippolytus’ prayer which would have been repudiated by Justin sixty years earlier.

How far, then, does the tradition represented by Hippolytus’ prayer go back? I shall suggest later that at least the general structure of the first part of Hippolytus’ prayer was an inheritance from the days of the Jewish apostles at Rome, which the Roman church with its usual conservatism had maintained more rigidly in the second century than some other churches. We shall find that this prayer as a whole is more ‘tidy5 in arrangement and more logical in its connections, less confused by the later introduction of inessentials, and more theological and precise in its expression of what is involved in the eucharistic action, than the others we shall consider. Here it is necessary only to draw attention to the careful articulation of its central portion (e-j).

The only point of any difficulty which arises in interpreting this prayer is the question of the exact bearing of (e). Is it to be understood as stating that our Lord went to His ‘voluntary passion’ in order’that He ‘might abolish death’ etc.; or does Hippolytus mean that He instituted the eucharist in order that ‘He might abolish death,’ etc.? Grammatically the sentence could mean either; and though to our way of thinking the former meaning may seem much more obvious, it seems from other passages in Hippolytus’ works that he did think of holy communion precisely as the means whereby Christ intended to bestow on us these benefits of His passion. Thus he speaks of communion as ‘the food which leads thee back to heaven, and delivers from the evil powers and frees from hard toil and bestows on thee a happy and blessed return to God.’ Similarly, commenting on Luke xxii. 15 (‘With desire have I desired to eat this passover with you before I suffer’) Hippolytus remarks, ‘This was the passover which Jesus desired to suffer for us. By suffering He released from sufferings (cf. Prayer (d) above) and overcame death by death and by a visible food bestowed on us His eternal life.... Therefore He desired not so much to eat as He desired to suffer that He might deliver us from suffering by (our) eating’ In the face of these and certain other expressions which Hippolytus uses elsewhere, it seems unnecessary to argue further. Hippolytus regards holy communion as the means by which Christ ‘abolishes death’ and ‘rends the bonds of the devil’ in the faithful communicant. It is a means of ‘enlightenment’ and a ‘demonstration of the resurrection’ (cf. John vi:53-57). The institution at the last supper ‘establishes an ordinance’ — a phrase in itself difficult to interpret of the passion.

The institution narrative of (f) is in fact the pivot of the whole prayer as it stands. It is the climax or point of all that precedes, and the starting point of all that follows. The command and promise it contains (g) are the justification for all that is done and meant by the church at the eucharist. This is carefully defined in (h), (i), (j), as (1) the offering of the bread and cup (2) which is the ‘priestly’ action of the church, and therefore a sacrifice (3) because it is the anamnesis of His own death and resurrection commanded by our Lord to be ‘done’; or as Justin (sup.) calls it, ‘What Jesus Christ our Lord commanded to be done for an anamnesis of His passion, which He suffered on behalf of men whose souls have (thereby) been cleansed from all iniquity.’ In other words, the eucharist was regarded in the second century as the divinely ordered ‘anamnesis’ of the redeeming action of our Lord. A good deal therefore turns on the word anamnesis, which we have so far left untranslated.

This word, which the Authorised Version translates as ‘Do this in remembrance of Me’ in the New Testament accounts of the institution, is more common in Roman writers in connection with the eucharist than elsewhere in pre-Nicene times. As we shall see, it does not appear in the parallel sections of some traditions of the prayer. It is not quite easy to represent accurately in English, words like ‘remembrance’ or ‘memorial’ having for us a connotation of something itself absent, which is only mentally recollected. But in the scriptures both of the Old and New Testament, anamnesis and the cognate verb have the sense of ‘re-calling’ or ‘re-presenting’ before God an event in the past, so that it becomes here and now operative by its effects. Thus the sacrifice of a wife accused of adultery (Num. V:15) is ‘an offering "re-calling" her sin to (God’s) remembrance’ (anamimnes-kousa); i.e. if she has sinned in the past, it will now be revealed by the ordeal, because her sin has been actively ‘re-called’ or cre-presented’ before God by her sacrifice. So the widow of Sarepta (1 Kings xvii:18) complains that Elijah has come ‘to "re-call" to (God’s) remembrance (anamnesai) my iniquity,’ and therefore her son has now died. So in Heb. x:3-4, the writer says that because ‘it is not possible that the blood of bulls and goats should take away sins’ (in the sight of God), the sacrifices of the Old Testament were no better than a ‘re-calling’ (anamnesis) of the offerers’ sins before God. And though in this passage there is some indication that anamnesis has here partly at least a psychological reference to the Israelites’ own ‘conscience’ of sins, it is plain from the passage as a whole that it is primarily before God that the sins are ‘re-called’ and ‘not purged’ or ‘taken away.’ It is in this active sense, therefore, of ‘re-calling’ or ‘representing’ before God the sacrifice of Christ, and thus making it here and now operative by its effects in the communicants, that the eucharist is regarded both by the New Testament and by second century writers as the anamnesis of the passion, or of the passion and resurrection combined. It is for this reason that Justin and Hippolytus and later writers after them speak so directly and vividly of the eucharist in the present bestowing on the communicants those effects of redemption — immortality, eternal life, forgiveness of sins, deliverance from the power of the devil and so on — which we usually attribute more directly to the sacrifice of Christ viewed as a single historical event in the past. One has only to examine their unfamiliar language closely to recognise how completely they identify the offering of the eucharist by the church with the offering of Himself by our Lord, not by way of a repetition, but as a ‘re-presentation’ (anamnesis) of the same offering by the church ‘which is His Body.’ As S. Cyprian puts it tersely but decisively in the third century. ‘The passion is the Lord’s sacrifice, which we offer’ (Ep. 63:17).

These three points may be said to stand out from our cursory examination of the Roman eucharistic prayer: (i) The centrality in its construction of the narrative of the institution as the authority for what the church does in the eucharist. Its importance in this respect is greatly emphasised by being placed out of its historical order, after the thanksgiving for the passion. (2) What is understood to be ‘done’ in the eucharist is the church’s offering and reception of the bread and the cup, identified with the Lord’s Body and Blood by the institution. This ‘doing’ of the eucharist is our Lord’s command and a ‘priestly’ act of the church. (3) The whole rite ‘recalls’ or ‘re-presents’ before God not the last supper, but the sacrifice of Christ in His death and resurrection; and it makes this ‘present’ and operative by its effects in the communicants.

We have no pre-Nicene text of the eucharistic prayer from Egypt. The earliest document of this tradition which has come down to us is a prayer which is ascribed in the unique eleventh century MS. to S. Sarapion, bishop of Thmuis in the Nile delta from before a.d. 339 to some date between a.d. 353 and c. a.d. 360. Whether the ascription to Sarapion personally be correct or not (and it is quite possible, despite certain difficulties) the prayer is undoubtedly Egyptian, and in its present form of the fourth century, from before rather than after c, a.d. 350. But there are strong indications that this extant form is only a revision of an older Egyptian prayer, whose outline can be established in some points by comparison with eucharistic passages in third century Egyptian writers. We shall not go into this reconstruction in any detail here. Our business is only to establish summarily certain differences from the third century Roman prayer of Hippolytus, and also certain very important similarities of ideas, which seem to belong to the third century Egyptian basis underlying the present text, as well as to the present text itself.

(a1) It is meet and right to praise, to hymn, to glorify Thee, O uncreated Father of the Only-begotten Jesus Christ. We praise Thee, O uncreated God, Who art unsearchable, ineffable, incomprehensible by any created substance. We praise Thee Who art known of Thy Son the Only-begotten, Who through Him art spoken and interpreted and made known to every created being. We praise Thee Who knowest the Son and revealest to the saints the doctrines concerning Him: Who art known of Thy begotten Word and art brought to the sight and understanding of the saints (through Him).

(a2) We praise Thee, O Father invisible, giver of immortality. Thou art the source of life, the source of light, the source of all grace and truth, O lover of men, O lover of the poor, Who art reconciled to all and drawest all things to Thyself by the advent (epidemia) of Thy beloved Son. We beseech Thee, make us living men; give us a spirit of light, that we may know Thee, the true (God) and Him Whom Thou hast sent, Jesus Christ; give us (the) Holy Spirit that we may be able to speak and tell forth Thine unspeakable mysteries. May the Lord Jesus speak in us and (the) Holy Spirit and hymn Thee through us.

(b1) [For Thou art far above all principality and power and rule and dominion and every name that is named, not only in this world but also in that which is to come. Beside Thee stand thousand thousands and ten thousand times ten thousands of angels, archangels, thrones, dominations, principalities, powers: by Thee stand the two most honourable .six-winged Seraphim, with two wings covering the Face and with two the Feet and with two flying, and crying ‘Holy’; with whom receive also our cry of ‘Holy’ as we say

(b2) Holy, holy, holy, Lord of Sabaoth; full is the heaven and the earth of Thy glory.

(c) Full is the heaven, full also is the earth of Thine excellent glory. Lord of powers, fill also this sacrifice with Thy power and Thy partaking: For to Thee have we offered this living sacrifice, this unbloody oblation.]

(d1) To Thee have we offered this bread, the likeness of the Body of the Only-begotten. This bread is the likeness of the holy Body, because the Lord Jesus Christ in the night in which He was betrayed took bread and brake and gave to His disciples saying: Take ye and eat, this is My Body which is being broken for you for the remission of sins. Wherefore We also making the likeness of the death have offered the bread, and beseech Thee through this sacrifice to be reconciled to all of us and to be merciful, O God of truth;

(d2) [and as this bread had been scattered on the top of the mountains and gathered together came to be one, so also gather Thy holy church out of every nation and country and every city and village and house and make one living catholic church.]

(d3) We have offered also the cup, the likeness of the Blood, because the Lord Jesus Christ taking a cup after supper, said to His own disciples: Take ye, drink; this is the New Covenant, which is My Blood, which is being shed for you for remission of sins. Wherefore we have also offered the cup, offering a likeness of the Blood.

(e l) O God of truth, let Thy holy Word come upon (epidemesato) this bread that the bread may become Body of the Word, and upon this cup that the cup may become Blood of the Truth;

(e2) and make all who partake to receive a medicine (lit. drug) of life, for the healing of every sickness and for strengthening of all advancement and virtue, not for condemnation, O God of truth, and not for censure and reproach.

(f) For we have called upon Thy Name, O Uncreated, through the Only-begotten in (the) Holy Spirit.

(g) [Let this people receive mercy, let it be counted worthy of advancement, let angels be sent forth as companions to the people for bringing to naught of the evil one and for the establishment of the church.

(h) We entreat also on behalf of all who have fallen asleep, of whom also this is the ‘re-calling’ (anamnesis) — (There follows the recital of the names) — sanctify these souls, for Thou knowest them all; sanctify all who have fallen asleep in the Lord and number them with all Thy holy powers and give them a place and a mansion in Thy kingdom.

(i) And receive also the eucharist of the people and bless them that have offered the oblations (prosphora) and the eucharists, and grant health and soundness and cheerfulness and all advancement of soul and body to this whole people.]

(j) Through Thy Only-begotten Jesus Christ in <the> Holy Spirit: (R/ of the congregation) As it was and is and shall be unto generations of generations and world without end. Amen.

This is much longer than Hippolytus’ prayer, but from the point of view simply of eucharistic teaching it says no more than the terse and direct theological statements of the Roman prayer, and it says it less precisely and adequately. A variety of new themes have found their way into the contents, but they obscure the simple outline found in Hippolytus without adding anything essential to the scope. The structure may be analyzed thus:

(a) Address. This is much more elaborate than that of Hippolytus, but is concerned with the same subject, the relation of God the Father to God the Son (to the exclusion in each case of the Holy Ghost). The first paragraph directly repudiates the teaching of Arius that the Son does not know the essence of the Father and is a creature. This makes it clear that it has been re-written (or perhaps added bodily before the second paragraph) during the second quarter of the fourth century, when the Arian controversy was at its height. If the older formula contained anything equivalent to Hippolytus’ thanksgivings for creation, incarnation and passion, only the faintest traces remain, in the references to ‘every created being’ and ‘the advent’ of the Son, with no allusion to the passion at all.

(b) Preface. What seems to have altered the character of (a) is the introduction of the sanctus, and of the preface introducing it. The note of ‘thanksgiving’ and the word itself have disappeared from the address, which has become a sort of theological hymn leading up to the preface. Omitting certain very interesting theological changes in (b) which can be shown to have been made in the fourth century, we note only that the use of the sanctus at the Alexandrian eucharist, preceded by a preface closely resembling Sarapion (b), can be traced in the writings of Origen at Alexandria c. a.d. 230 (Ibid). This is the earliest certain evidence of the use of this hymn in the liturgy. Earlier citations of the words of the angelic hymn from the scriptures by Clement of Rome and Tertullian do not necessarily reflect a use of it at the eucharist, and it is absent from Hippolytus’ liturgy and from some other early documents. It is also noticeable that while the later Alexandrian Liturgy of S. Mark shews little trace in other parts of its eucharistic prayer of being descended from a prayer at all closely resembling that of Sarapion, in the one point of the wording of its preface S. Mark exhibits only small verbal variations from the text of Sarapion (b). The simplest explanation of these various facts is that the use of the preface and sanctus in the eucharistic prayer began in the Alexandrian church at some time before a.d. 230, and from there spread first to other Egyptian churches, and ultimately all over Christendom. If this be true, Sarapion’s (b), though an integral part of the text in its present (fourth century) form, is an interpolation into the original local tradition of the prayer at Thmuis, as is indicated by its having been borrowed almost verbally from the liturgy of Alexandria. We have no means of judging when this Alexandrian paragraph was first incorporated into the liturgy at Thmuis, whether as part of that revision which formed our present text of the prayer — which is certainly responsible for the present form of (a) and may quite well have included a recasting of the whole opening part of the prayer (Sarapion was a close friend and prominent supporter of S. Athanasius, bishop of Alexandria from a.d. 328-373) — or by some earlier revision at Thmuis during the third century. But at Thmuis the preface has received no local development or variation worth mentioning from the Alexandrian text, which in the conditions of the period suggests that its incorporation was not of long standing when the present revision was made.

(c) Prayer for the acceptance of the ‘living sacrifice.’ This section is difficult to interpret. At first sight it marks an abrupt transition from the worship of the sanctus to the offering of the eucharistic oblation of the bread and the cup. The phrase ‘the unbloody sacrifice’ is used by fourth century writers (first by Cyril of Jerusalem a.d. 348) to mean the specifically eucharistic offering of the consecrated bread and cup; and a prayer having a definite reference to the consecration of the bread and cup, at this point before the recital of the institution, is a peculiar characteristic of some later Egyptian eucharistic prayers.

Nevertheless it is open to doubt whether this was the original application of (c), even if by Sarapion’s time it had already come to be interpreted in this sense. There is a certain difficulty in the prayer that God would ‘fill this sacrifice’ with His ‘partaking,’ which is awkward on any interpretation, but especially so if (c) be really a prayer about the bread and the cup. And there is an unexpectedness about the phrase ‘this living sacrifice’ applied to the elements on the altar at this stage of the prayer without any sort of warning, even allowing for the fact that the idea of a ‘moment of consecration’ had hardly developed in the fourth century (as the next section of the prayer sufficiently indicates). But it would be a good deal easier to understand if it has a connection with the previous petition, ‘we beseech Thee make us living men.’ In this case the ‘living sacrifice and unbloody oblation’ of (c) will have reference to the ‘sacrifice of praise’ offered in the hymn of the sanctus, and not to the eucharistic offering which follows. It is at least worthy of notice that in a pre-christian Jewish work (c. 100 B.C.) The Testament of the xii Patriarchs, the angels in heaven are said to offer ‘a rational and unbloody oblation’ to God, and it is in this angelic worship of heaven that the congregation has just been joining by the sanctus. Similarly a second century Christian writer, Athenagoras, speaks of’the lifting up of holy hands’ by Christians as ‘an unbloody sacrifice and rational liturgy,’ clearly with reference to prayer and praise rather than to the eucharist as such. In this case Sarapion (c) would represent originally a prayer for the acceptance of the sacrifice of praise offered in (b), much as (d1) contains a prayer for the acceptance of the eucharistic sacrifice of the bread and wine offered in the preceding sentence; and as (1) is a prayer for the acceptance of ‘the eucharist of the people’ offered in the whole preceding prayer.

Such an interpretation of (c) eases the abruptness of the main transition of thought, which comes not between (b) and (c), but between (c) and (d). The transitions are not very well managed anywhere in this prayer, but it seems easier at this point if there is a passage of ideas from the offering of the worship of the sanctus as a ‘living sacrifice’ of praise, to the offering of the eucharistic ‘sacrifice of the death.’ This carries with it the implication that (c) (which thus depends on the sanctus) is also an interpolation into the original form of the rite of Thmuis. A good deal has been built on the application of (c) in this prayer to the eucharist by some writers; but it does not really seem to make much difference to the specifically eucharistic theology of the prayer to exclude (c) from consideration in this respect.

(d) The Offering and Institution. As a preliminary to understanding this section it is best to dispose of (i2), which completely destroys the symmetry, otherwise obvious, between (d1) and (d3). The unsuitability of describing the corn from which the eucharistic bread has been made as having been originally ‘scattered on the tops of the mountains’ among the mud-flats of the Nile delta makes it plain that this is not an authentic product of the native tradition of the prayer at Thmuis, but a rather unimaginative literary quotation. It is in fact borrowed from the prayers for the agape found in Didache ix (Cf. p. 90). (In the Syrian or Transjordanian setting in which the Didache was probably composed, cornfields on the hill-tops occasion no surprise.) As an elaboration of (i1), (d2) is still a rather glaring ‘patch,’ which has not yet produced a similar elaboration of (j3). This suggests that it had not very long found a place in the prayer when the present recension was made. It may even have been introduced as a ‘happy thought’ by the last reviser, since it virtually duplicates matter found more in place in (g), which is itself an addition to the original outline of the prayer.

By contrast with Hippolytus, Sarapion in (d) fuses the formal statement of the offering of the elements with the narrative of the institution, which Hippolytus keeps distinct (cf. Hipp. (f) and (i)). Sarapion also states explicitly that the actual offering has already been made at the offertory, which Hippolytus leaves in the background. We have already seen the reason for this in the fact that ‘the’ prayer had originally to put into words the meaning of the whole rite, of what precedes as well as of what follows. Thus Sarapion can say ‘We have offered’ (before the prayer began) even though the whole prayer is itself headed in the MS. ‘Prayer of Offering’ or ‘Oblation.’ Finally, even more plainly than in Hippolytus, the narrative of the institution is here pivotal for the whole prayer, as the supreme authority or justification for what the church dees in the eucharist — ‘This bread is the likeness of the holy Body because the Lord Jesus took bread,’ etc.

(e) Prayer for Communion. This section forms a single whole, even though it falls into two distinct parts. It is a prayer for communion, the first part of which is concerned with the means and the other with the effects. In contrast with Hippolytus, where the institution narrative is taken as implicitly identifying the bread and wine with the Body and Blood of Christ by virtue of His own promise, Sarapion’s prayer shews a new desire for an explicit identification. This desire is found in other fourth century writers also, but hardly before that time. The way in which, e.g., (d3) goes out of its way to emphasise this identification of the bread and wine with the Body and Blood by the institution narrative itself, with the peculiar formula..’. drink, this is the New Covenant, which is My Blood’ (instead of ‘in My Blood,’ Luke 22:20), suggests that at one time the Hippolytan understanding of the force of the institution narrative had prevailed in Egypt also. It was only later that it was felt to need reinforcing by an explicit petition for the identification of the elements with the Body and Blood, such as we get here in (e).

However this may be, Sarapion is not unique in the fourth century in feeling this, or in the way in which he expresses himself, by a prayer for the ‘advent’ (epidemesato) of the Word, parallel to His cadvent’ (epidemia) in the incarnation (cf. a2). S. Athanasius in the same period in Egypt writes: ‘When the great prayers and holy supplications have been sent up the Word comes upon the bread and the cup and they become His Body.’ The same idea is found in a number of Ethiopic rites which are of Egyptian connection, if not actual origin. Outside Egypt S. Jerome in Syria sixty years later speaks of bishops as those who ‘at the eucharist pray for the advent of the Lord,’ and similar language is used in Asia Minor in the fourth century, and later still in Italy, Gaul and Spain. This introduction of a prayer for ‘the coming of the Lord,’ the Son, the Second Person of the Trinity, is a straightforward conception, which only makes explicit the ideas originally involved in the reference to the incarnation and in the institution narrative in earlier versions of the prayer. The implications of these references had already been made plain by writers like Justin in the second century. But the introduction of such a petition alters to some extent the balance of the prayer as a whole, by weakening the position of the institution narrative as the central pivot of the whole prayer.

Even in so early a specimen as that of Sarapion, the prayer of (e1) is definitely ‘consecratory’ in form, and thus prepares the way for the conception of a ‘moment of consecration’ within the eucharistic prayer as a whole. This conception was eventually accepted by East and West alike, though they chose different ‘moments’ to which to attach the idea. It was by a third development, a sort of theological refinement upon this secondary stage of any sort of explicit prayer to reinforce the old identification of the elements with the Body and Blood through the institution narrative, that the Greeks evolved during the fourth and fifth centuries the ‘tertiary’ stage of a prayer that specifically the Holy Ghost, the Third Person of the Holy Trinity, would (in some sense) ‘make’ the elements into the Body and Blood of Christ. This became for them the ‘moment of consecration’; a ‘moment’ which the West, when it adopted the idea from the East, continued to place at the old pivot of the prayer, the institution narrative. Had the West wished to follow the East in divorcing the ‘moment’ from the institution, it could have found one at the prayer Quam oblationem of the Western canon before the institution narrative, which is just as much ‘consecratory’ as is (el) in Sarapion. Rome therefore reached this secondary stage of a petition for consecration apart from the institution; but remained there, without advancing to the ‘tertiary’ stage of the Eastern prayer for the sending of the Holy Spirit. Sarapion’s prayer in (e1) thus foreshadows the parting of the ways between later Eastern and Western liturgical ideas. (e2) Having prayed for the means of communion, Sarapion prays for its effects. Here it is noticeable that whereas Hippolytus’ prayer for the communicants confines itself to purely spiritual effects, that of Sarapion recognises that the sacrament is a ‘drug’ or ‘medicine’ of life, for the body as well as the soul. We need not suspect that this difference represents a ‘rapid decline of spirituality between the days of persecution and those of the established church of the fourth century,’ as one English writer has suggested. (Sarapion himself felt the full force of the Arian persecution of the catholics, and probably died in exile.) It is quite true that Hippolytus at this point says nothing of the eucharist as concerned with the human body; but in his section (e) he has quite clearly stated that one purpose of the institution of the eucharist is ‘to abolish death’ etc., which amounts to much the same thing, though put in a different way. In point of fact, Sarapion rests on old Egyptian tradition in what he calls the eucharist here. Clement of Alexandria, c. a.d. 190, had pictured our Lord as saying to the soul: ‘I am thy nourisher, giving Myself as bread, whereof he that tastes shall never more have experience of death, and daily giving Myself for the drink of immortality.’ We shall see in the next chapter that these ideas go back right through the second century into the New Testament itself. The Roman canon follows the tradition of Hippolytus in that it prays only for spiritual benefits for the communicants — that ‘they may be filled with all heavenly benediction and grace,’ a conservatism which is followed by our Prayer Book ‘Prayer of Oblation.’ But our words of administration — ‘preserve thy body and soul’ — have gone back to the wider view of the effects of communion, by contrast with the Roman words — ‘preserve thy soul unto everlasting life.’ In more discreet language our form contains Sarapion’s teaching that the eucharist is a ‘drug’ or ‘medicine of life’ for the body as well as the soul.

(f) The Invocation. We have already spoken of the great importance attached in the primitive Christian and the pre-christian Jewish tradition to the ‘glorifying of the Name’ of God at the close of the berakah or eucharistia, the ‘Thanksgiving’ at the end of supper. We have a further hint in this clause of the part played by this conception. The prayer in (el) and (e2) for the identification of the elements with the Body and Blood of Christ and for their eternal effects upon the bodies and souls of the communicants — the petition of the whole eucharistic prayer — is here understood as being efficacious chiefly ‘because they have called upon the Name of God.’ So again, Clement of Alexandria, citing an even earlier Egyptian writer c. a.d. 160, with whom Clement does not disagree on this point, says: The bread is hallowed by the power of the Name of God, remaining the same in appearance as it was (when it was) taken, but by (this) power it is transformed into spiritual power.’

Whatever the danger of approximating to mere magic in such ideas, we have to recognize that the special efficacy of prayer ‘in the Name of God’ or ‘of the Lord Jesus’ is clearly found in the New Testament, not only in the teaching of the apostles — and in their practice, e.g., in the matter of exorcisms — but also in the teaching of our Lord Himself. There is no clear dividing line to be drawn between the application of such ideas to the sacrament of the eucharist, and to that of baptism, whether this be given ‘in the Name of the Holy Trinity or, as primitively, ‘in the Name of the Lord Jesus.’ We accept it placidly in the case of baptism out of use and wont, because the church happens to have retained it in its full primitive significance in baptism. We are startled at it in the case of the eucharist, because there the church early overlaid it with other ideas. But in the time of Sarapion it had not yet entirely lost its primitive force in the eucharist, and it is likely that this clause was deliberately retained out of a lingering sense of the importance of the old conception, when the intercessions which follow in the present (fourth century) text were first interpolated at this point in the prayer.

(g), (h), (i) The Intercessions, for the Living, the Dead and the Offerers. These are an addition to the original outline of the prayer, of a kind which was made in most churches at some point within the prayer before the end of the fourth century. When the eucharist was celebrated apart from the synaxis in the pre-Nicene church there was a real loss in the absence of any intercessions whatever. There was a natural desire to replace them in some way; and it is quite possible that in some churches the custom arose during the third century of treating the intercessory ‘prayers of the faithful,’ which really formed the close of the synaxis, as a sort of invariable preliminary to the eucharist, even when this latter was celebrated without the rest of the synaxis. (But Sarapion’s own arrangement in his collection of prayers still puts the intercessions at the opposite end of the book to the prayers of the eucharist proper, in an altogether separate service.)

The alternative was to insert some intercessions at a fresh point within the eucharist itself. The rigidity of the primitive outline, which permitted of only one prayer at the eucharist, ‘the’ eucharistic prayer, necessitated their being included somehow within that, whatever confusion to its primitive shape and purpose this might cause. Even when the two services were celebrated together, there was a natural desire to associate a prayer for the ‘special intentions’ with which the eucharist was being offered as closely as possible with the act of offering, and this would lead to the same result. The existence of some prayer for the communicants towards the close of the prayer (in all the traditions with which we are acquainted) led in some churches to the development of this part of the prayer to cover other objects of intercession as well, as here at Thmuis, and also at Jerusalem, where it is probable that the practice started. In the fourth century such a position for intercessions acquired the further sanction of the idea of the special efficacy of prayer in the presence of the consecrated sacrament, which we shall find attested by S. Cyril of Jerusalem in a.d. 348 (Cf.p. 199). But Jerusalem in the fourth century, and especially S. Cyril, are in the forefront of ‘liturgical advance,’ and there is no sign of this further special development of ideas in Sarapion.

Alexandria and Egypt generally adopted another notion, that the special intentions of the sacrifice ought to be named before it was actually offered. We find accordingly that the Alexandrian intercessions were inserted into the opening of the prayer, before the sanctus. At Rome the intercessions for the living settled down at the beginning of the prayer (but after the sanctus), and those for the dead (originally only inserted at masses for the dead) at the end. Elsewhere other points were chosen; e.g., at Edessa they were interpolated after the sanctus and the first half of the eucharistic prayer, immediately before the consecration (Cf. p. 179 n). There was no uniformity about this, because each church began to copy others in ‘modernizing’ its liturgy at different moments and under different influences, inserting now the preface and sanctus, now intercessions for the living, now commemorations of martyrs and so on, at whatever point in its own local tradition of the prayer seemed most fitting; and in doing so it borrowed now verbally, now only in ideas, now from one source, now from another, or added native compositions and elaborations of its own as the liturgical gifts and knowledge of its successive bishops permitted.

The general result, when the synaxis and eucharist came to be fused into a single rite, celebrated as a normal rule without a break, was a duplication between the old intercessions, the ‘prayers of the faithfiiT, at the close of the synaxis, and the new intercessory developments within the eucharistic prayer. The old ‘prayers of the faithful’ tended after a while to atrophy in most rites, or even to disappear’altogether, as at Rome and in the Syriac S.James.

The chief points of interest in Sarapion’s intercessions are: (h) The description of the eucharist as the anamnesis of the dead — clearly in the same sense as at Rome of ‘re-calling’ something before God. But the word is not applied to the eucharist as the anamnesis of the passion in Sarapion, though it is found in this sense in Origen in third century Egypt. In (i) the prayers for the offerers are of interest as the earliest Egyptian evidence for the custom of each communicant bringing his or her own prosphora for themselves. To be one of ‘the people’ (laity), to offer the prosphora and to partake of communion, were still all virtually the same thing in Sarapion’s time in Egypt, to judge by the way the petitions in (e2), (g), and (i) repeat one another in their prayers for Advancement.’ In the later Alexandrian intercessions also, those for the dead immediately precede those for the ‘offerers.’

(j) The Doxology. In the present text this is reduced to meagre dimensions. Probably the interpolation of the intercessions has eliminated an older fully developed form at (f), which marked the conclusion of the prayer. That (j) does not preserve the original conclusion postponed to the end of the interpolated intercessions, seems clear from the fact that the traditional people’s response ‘As it was’ etc., does not attach itself to Sarapion’s conclusion either grammatically or in sense, though it is appended in the MS.

Comparing the whole prayer with that of Hippolytus one may say that though it is more than probable that Sarapion ultimately derives from a prayer on the berakah model, and though there are certain points of contact between Hippolytus and Sarapion in structure, it has in any case lost touch with its original type much more than has the older Roman prayer. Additional themes like the sanctus and the intercessions have complicated and obscured the outline so much that no clear verdict could be given on this question of derivation from the berakah from the study of Sarapion’s prayer taken alone. And certainly there has been no borrowing between the Roman and Egyptian prayers in the course of development. In the central part of the prayer [Sarapion (d)-(j) = Hippolytus (e)-(j)] the differences of phrasing and arrangement are very marked indeed, considering that both prayers are dealing with exactly the same subject.

But this obvious independence of the two traditions only brings into greater relief their agreement on the substance of those points which we noted as outstanding in Hippolytus’ statement of the meaning of the eucharistic action:

(i) The bread and the cup are explicitly stated to be ‘offered’ to God — though in Sarapion separately, in Hippolytus together. (2) Sarapion explicitly calls this a ‘sacrifice,’ as Hippolytus calls it a ‘priestly’ ministry; the meaning is the same though the statement is diverse. Though the eucharist is not called ‘the anamnesis of the passion,’ as in Justin and Hippolytus, it is called ‘making the likeness of the death.’ And (3) as in Hippolytus, the pivotal importance of the narrative of the institution in the prayer, as the ground of the eucharist’s effective ‘re-calling’ before God of tie sacrifice of Christ, does not in any way obscure the fact that it is Calvary and not the Upper Room which is thus ‘re-called.’

In Syria the church of Antioch claimed and was accorded a primacy from, at the latest, some while before the end of the second century. But for a variety of reasons this was never so effectively exercised as was that of Alexandria over Egypt. Despite a cleavage of race and language between the native Copts and the large population of immigrant Greeks, Egypt had been a self-conscious unity under the leadership of Alexandria for centuries before the coming of Christianity. The unchallenged supremacy of the Alexandrian bishop over all the churches of Egypt only gave Christian expression to an enduring political and geographical factor in past Egyptian history. But from pre-historic times Syria has always been a mosaic of different races, cultures, religions and languages, which no political framework has ever held together for long. The welter of Canaanite tribes of very diverse racial origin which the Hebrews under Joshua succeeded in overcoming in the hills of Southern Syria is typical of the pre-historic background of the whole country. It is equally typical of its history that the invading Israelite confederacy should promptly have disintegrated into its original tribal units under the Judges; and even after it had been welded into a single state under Saul and the House of David, should have split again after less than a century into the rival states of Israel and Judah. The North and East of Syria were no less prone to division than the South throughout their history — until only yesterday, when the four separate republics of French Syria and the two states of Palestine and Transjordan under British mandate still divided a country which seems geographically destined to be a unity, but which is racially and culturally one of the least united in the world.

During the century c. 250-150 B.C., the Seleucid kings of Antioch made the most promising of all the many attempts to unify Syria, on the basis of the introduction everywhere of Greek language and culture. They hoped this would be a general solvent of all the diverse local traditions, and act as a cement for the motley elements over which they ruled. They were thwarted by the stubborn adherence of large parts of the population to their ancient cultures, of which the resistance of the jews of the South under the Maccabees is only the most obvious and violent example.

The Seleucids failed in their main object, but they had a good deal of incidental success with their chosen means, the introduction of that form of later Greek civilization which we call ‘Hellenism.’ Henceforward Syria was riven by a new division, running right across all its old fractions, that between hellenism and the old native cultures, which diverse though they were, may be classed together as predominantly Semitic. This new cleavage does not run along racial lines, for the vast majority of the hellenists were not immigrants but hellenised Syrians. Nor was it primarily geographical, though naturally Antioch and the great coast towns were strongholds of hellenism, as the hinterland was of the native tradition. But there were large purely oriental quarters in Antioch itself and whole Aramaic-speaking districts in its neighbourhood; on the other hand there were at times strong Greek influences at work in Edessa and Damascus, inland cities which were normally centres of Semitic culture; while some of the smaller cities on the Eastern frontier were completely hellenised. The backbone of the Semitic tradition was the peasantry of the countrysides, as the peak of hellenism was found in the towns. But there were Greek-speaking country districts, while some towns, especially in the East — Edessa, Palmyra, Damascus — were strongly Semitic by tradition, and others like Aleppo and Emesa (Horns) formed a sort of debatable land between the two cultures. In short, Syria was an older underlying patchwork of races, languages, traditions and religions, with a recent and different patchwork of hellenism and the surviving native cultures superimposed upon it. The underlying patchwork is local, but the only line of division one can draw between hellenism and the oriental traditions is purely cultural By a.d. 300 a man might be a Syrian (which could mean racially a mongrel of half-a-dozen different strains) and yet as hellenised and westernized in speech and mind and habit of life as an inhabitant of Athens or Alexandria or even Rome. And his next-door neighbour might be equally Syrian by blood and remain as completely oriental in culture and language and thought as his forefathers a thousand years before. Or he might be bilingual, with some sort of footing in both worlds. First Rome and then Byzantium inherited the hellenising policy of the Seleucids; and while these European powers ruled the land, Antioch, which had been founded as the capital of hellenism in Syria, remained the administrative and ecclesiastical capital. With the return of Semitic ascendancy after the Arab conquest in the seventh century, dominance returned to the old Semitic centre of Damascus, to which both the Arab rulers and the Christian patriarchs transferred their courts. Henceforward Antioch slowly declined into insignificance.

The patriarchate of Antioch saw itself as the Christian heir to the Seleucid tradition of the leadership of all Syria in the path of hellenism; and with only two brief exceptions (under the heretical patriarchs Paul of Samosata in the third century and Severus in the sixth), it identified itself with the ‘royalist’ hellenising movement throughout its history. But in adhering to this policy the patriarchs had to face in the ecclesiastical field just those same centrifugal tendencies and obstinate local traditions which faced every attempt at political centralization. When Bishop Juvenal of Jerusalem in a.d. 451 succeeded after twenty-five years of manoeuvring in extracting from the general council of Chalcedon formal recognition of his see as an independent patriarchate over Palestine, he only added a Christian chapter to the long story of the wars of Israel with Syria which punctuate the Books of the Kings, and are continued by the revolt of the Maccabees against the Seleucids. And besides this inveterate separatism of the South there were other pockets of local resistance to all Antiochene or hellenistic domination, less strongly marked but in the end equally tenacious. Against the overwhelming political power of Rome or Byzantium these local patriotisms could only express themselves in terms of ecclesiastical resistance, under the pretext of doctrinal heresy culminating in schism. But these dissident churches drew their strength from racial and cultural forces far more than from theological nicety. Apart from a whole succession of obscure and fantastic popular movements like that of the Messalians in the fourth century (most of which were hardly sufficiently Christian to be classed as heresies) we have to reckon, first, with the great East Syrian revolt against Antioch in the fifth century, which adopted the banner of the Nestorian heresy; and secondly, with its doctrinal opposite, the West Syrian revolt of the sixth century which called itself Monophysite; and thirdly, with the Maronite schism in the Lebanon of the eighth century, which took the excuse of Monothelitism. We need not here concern ourselves with the doctrinal pretexts. The real dogma of all the rebels was ‘anti-Byzantinism’ or ‘anti-hellenism’ as the ‘orthodoxy’ of Antioch was always in practice ‘Caesaro-papism.’ Between them the royalist patriarchate and the nationalist schisms shattered Syrian Christianity as a living force, and left it permanently weakened to face the pressure of moham-medan political conquest. To-day more than three quarters of the descendants of the old Christian inhabitants of Syria are mohammedans, and the Christian remainder is so riven into fragments as to be a negligible missionary power. The islamic populations of Syria and Egypt no less than their schismatic churches are permanent monuments of the long attempts of the church of Constantinople to dominate the Christian world in the interest of the Byzantine emperors.

It is not surprising that this background of abiding cultural division and local separatism should have left its mark on the liturgy. But the liturgical divisions of Syria, by a series of historical accidents, do not entirely coincide with those of ancient ecclesiastical politics or present doctrinal allegiance. In the field of liturgy we can distinguish four main influences which cross the present sectarian divisions in a most confusing way:

(1) The old rite of the church of Antioch itself, which is very imperfectly known;

(2) The other early West Syrian liturgical traditions, which we shall ignore;

(3) The East Syrian tradition, centered in Edessa;

(4) The South Syrian tradition of Jerusalem.

(1) What may be called the ‘patriarchal’ rite of Syria was the so-called Liturgy of S. James. It is generally taken that, as it stands, this is not the old local rite of Antioch, which is known to us only obscurely from a number of sources, of which the most reliable are hints to be found in the Antiochene writings of S. John Chrysostom (c. A.D. 360-397). S. James as it stands is closely connected with the fourth century rite of Jerusalem, which was adopted by the Antiochene church at some point in the fifth century — when is uncertain. It had not yet happened when S. John Chrysostom left Antioch in A.D. 397, and it is reasonable to suppose that it did not happen after a.d. 431, when Bishop Juvenal of Jerusalem greatly embittered relations between Jerusalem and Antioch by claiming not merely independence (which he successfully asserted twenty years later) but jurisdiction over Antioch itself for his own see. The unique position of Jerusalem as the ‘holy city’ and above all its prestige as a model of liturgical observance were such during the turn of the fourth and fifth centuries as to cause the adoption of Jerusalem customs to a greater or lesser extent by other churches all over Christendom. It is not surprising that it should have influenced its own patriarchal see in these respects with especial force at this time. At all events, Antioch to some extent adopted and adapted the Jerusalem Liturgy of S. James, probably between A.D. 400 and 430, and made it the patriarchal rite so far as Antiochene influence extended.

Strangely enough, though the patriarchs of Antioch thus introduced the Jerusalem rite into North Syria, they did not themselves remain faithful to it, and ultimately abandoned its use altogether. In pursuit of their usual hellenising policy they had begun (? in the seventh century) to use a version of the Greek Liturgy of S. Basil, as at least an occasional alternative to their own rite of S. James. After some centuries of increasing ‘Byzantinising,’ they ended in the thirteenth-fourteenth century by dropping all trace of their own Syrian rite in favour of the full rite of Byzantium, upon which power the Antiochene orthodox patriarchate had by then become helplessly dependent. Thus S. James, though the patriarchal rite of Antioch, is neither a ‘pure’ descendant of the original rite of the Antiochene church, nor the rite which has been used by its patriarchs for the greater part of their history.

(2) North-West Syria followed its patriarchs in adopting S. James, but with one important reservation. While the structure and framework of S. James everywhere came into use, the text of its eucharistic prayer never achieved the same prescriptive authority in N.W. Syria as the rest of the rite. Some seventy alternative eucharistic prayers are known from this region, composed at all periods from the fourth-fifth centuries down to the fifteenth. In other words, the working authority of the Antiochene patriarchate was never sufficiently strong in the nearest parts of its own territory, even before the great revolts of the sixth century, to break down the old tradition that every church could follow its own usage in the phrasing of its eucharistic prayer, and that celebrants could remodel this within certain limits at their own discretion. The general outline of these prayers follows that of S. James fairly closely as a rule. But some of them exhibit very interesting and probably ancient variations, and have been only roughly adapted to fit the S. James type; while even those prayers which follow it more closely are verbally independent compositions on the same theme rather than mere imitations.

But by the time of the Monophysite schism (sixth century) S. James had obviously become the standard West Syrian tradition. For a while after that royalists and schismatics used the same rite, until the royalists came to think of it as a badge of local particularism and abandoned it for the rite of Constantinople. This left it to the exclusive use of the Mono-physites, among whom it now survives in an Arabic translation, though before the seventeenth century it was generally used in an ancient Syriac version (which is still in use in a few Christian villages round Damascus). The Syriac appears to have undergone more than one revision since the sixth century, sometimes to bring it into greater conformity with Byzantine innovations, sometimes in complete independence of these. Even in their hostility to Byzantium the provincials could not help being more than a little impressed by the Byzantines’ own valuation of themselves as the source of all that was ‘correct’ in matters ecclesiastical. They were consequently always apt to adopt the latest Byzantine customs after more or less delay, and so gradually to Byzantinise their own rites. Modern and mediaeval Monophysite MSS. of S. James differ textually from one another more considerably than those of any other rite — another symptom of the permanent lack of central authority in matters liturgical in Syria.

(3) North-East Syria seems never to have adopted S. James having gone off into Nestorianism and independence too early to have been much influenced by its adoption by the patriarchs of Antioch. Instead, this part of the country adopted as its standard liturgy the ancient rite of the church of Edessa, the Liturgy of SS. Addai and Mari (the traditional ‘apostles’ of Edessa). This may well be connected originally with the second century rite of Antioch, whence Edessa had received the faith; though this is no more than a very reasonable conjecture. Edessa was a semi-independent state on the Eastern Roman frontier, a strong centre of Semitic culture and tradition, though theologically it also acted as a channel for the diffusion of Greek ideas to the purely unhellenic regions around and east of itself. Even Nestorius, whose teachings the later school of Edessa professed to follow, was an ecclesiastic of Antioch who became patriarch of Constantinople; and his teachers Theodore and Diodore, who were venerated as Nestorian ‘doctors,’ were likewise thoroughly hellenised, even though all three were from inner Syria and probably racially non-hellenic. The Edessan liturgy has therefore undergone some infiltration of hellenic ideas even in the earliest texts now available.

But it is of unique interest and importance none the less, because it is basically still a Semitic liturgy, the only remaining specimen of its kind. It is cast in a different idiom of thought from that of the eucharistic prayers of the hellenistic Christianity which had developed out of S. Paul’s missions to the hellenistic world north and west of Syria. Its special importance lies in this — that any agreement of ideas with these hellenistic prayers which may be found to underlie the marked peculiarities of SS. Addai and Mari helps to carry back the eucharistic tradition of the church as a whole behind the divergence of Greek and Western Christianity generally from that oriental world to which the original Galilaean apostles had belonged. The obscure history of the Syrian liturgies has a special interest just because it illustrates that contrast between the whole mind and thought of the hellenic and Semitic worlds which rarely meets us with any definiteness in Christian history outside the pages of the New Testament. We shall therefore conclude this chapter by examining two Syrian eucharistic prayers which are expressions of the two aspects of Syrian tradition, those of the more Semitic Liturgy of SS. Addai and Mari and of the more hellenistic Liturgy of S. James. There is much to be learnt from their different ways of expressing what is fundamentally the same liturgical tradition.

(a) Worthy of praise from every mouth and of confession from every tongue and of worship and exaltation from every creature is the adorable and glorious Name [of Thy glorious Trinity, O Father and Son and Holy Ghost,]

(b) Who didst create the world by Thy grace and its inhabitants by Thy mercy and didst save mankind by Thy compassion and give great grace unto mortals.

(c1) [Thy majesty, O my Lord, thousand thousands of those on high bow down and worship, and ten thousand times ten thousand holy angels and hosts of spiritual beings, ministers of fire and spirit, praise Thy Name with holy Cherubim and spiritual Seraphim offering worship to Thy sovereignty, shouting and praising without ceasing and crying one to another and saying:

(c2) Holy, holy, holy Lord God of Hosts; heaven and earth arc full of His praises and of the nature of His being and of the excellency of His glorious splendour. Hosanna in the highest, and Hosanna to the Son of David! Blessed is He that came and cometh in the Name of the Lord! Hosanna in the highest! And with these heavenly hosts]

(d) We give thanks to Thee, O my Lord, even we Thy servants weak and frail and miserable, for that Thou hast given us great grace past recompense in that Thou didst put on our manhood that Thou mightest quicken it by Thy Godhead,

(e) and hast exalted our low estate and restored our fall and raised our mortality and forgiven our trespasses and justified our sinfulness and enlightened our knowledge, and, O our Lord and our God, hast condemned our enemies and granted victory to the weakness of our frail nature in the overflowing mercies of Thy grace.

(/) And we also, O my Lord, Thy weak and frail and miserable servants who are gathered together in Thy Name, both stand before Thee at this time

(g) and have received by tradition the example which is from Thee,

(h) [rejoicing and glorifying and exalting and commemorating and performing this (great and fearful and holy and life-giving and divine) likeness of the passion and death and burial and resurrection of our Lord and our Saviour Jesus Christ.]

(i) And may there come, O my Lord, Thy Holy Spirit and rest upon this oblation of Thy servants, and bless and hallow it that it be to us, O my Lord, for the pardon of offences and the remission of sins and for the great hope of resurrection from the dead and for new life in the kingdom of heaven with all those who have been well-pleasing in Thy sight.

(j) And for all this great and marvelous dispensation towards us we will give Thee thanks and praise Thee without ceasing in Thy church redeemed by the precious Blood [of Thy Christ], with unclosed mouths and open faces lifting up praise and honour and confession and worship to Thy living and life-giving Name now and ever and world without end.

R/ Amen.

Before commenting in detail on this prayer there are two general observations of some importance to be made, (i) So far as can be ascertained the biblical text which underlies the scriptural citations in this prayer is not a Greek text, but one of the Syriac versions — which, it is not possible to distinguish. It would appear certain, therefore, that unlike most other Eastern vernacular rites, Addai and Mari was not originally a translation from the Greek, but was composed in Syriac.

(2) Whatever may be the case in the opening address of the prayer and certain phrases elsewhere, the body of this eucharistic prayer is undoubtedly addressed not to the Father but to the Son. Phrases such as ‘Thou didst put on our manhood’ (d), and ‘the example which is from Thee’ (f), are quite inapplicable to the First Person of the Trinity; and ‘Thy ... servants who are gathered together in Thy Name’ is a reference to Matt. 28:20 — ‘Where two or three are gathered together in My (our Lord’s) Name, there am I in the midst of them.’ However surprising the idea of a eucharistic prayer to the Son may seem to us, it was not very unusual in antiquity. Besides the Egyptian Liturgy of S. Gregory and another Egyptian eucharistic prayer published by Hyvernat, there are three Ethiopic liturgies all addressed to the Son. In Syria itself the Monophysite Second Liturgy of S. Peter and two lesser Maronite liturgies are directed to the Son, as is part of the eucharistic prayer of the Syriac S. James itself (Cf.p. 190 n.), which is followed in this by nearly all the sixty or seventy lesser Syriac liturgies. Evidently there was a strong tradition on this point in Syria generally. In the West there are distinct traces of such a custom having once been common in Mozarabic and Gallican eucharistic prayers; and the repeated condemnation of the practice by two North African councils at the end of the fourth century proves that it was not unknown there either. The fact that SS. Addai and Mari is addressed to the Son is thus only a proof of antiquity, and not an exceptional peculiarity.

(a-c) Address, Memorial of Creation, Preface and Sanctus. It seems fairly clear that the preface and sanctus, which have no connection with what precedes and follows, are an interpolation, and that Addai and Mari (like Hippolytus) originally did not contain any such feature. ‘Came and cometh’ in the Benedictus is found also in the Syriac S. James, which may give us a clue as to whence the whole passage was borrowed (cf. p. 188). What is more difficult to decide is the authenticity of (a) and (b). The address to the Trinity has obviously been rewritten, but Mr. Ratcliff has pointed out that (a) ‘Worthy ... of confession from every tongue ... is the Name .. . of Thy ... Trinity’ is reminiscent of Philippians ii. 9-11, where, however, ‘the Name’ is the Name of Christ. It seems, therefore, probable that the interpolation of the sanctus has led to the re-writing of (a) in Addai and Mari (much as we saw that it has done in Sarapion); but in Addai and Mari this has been effected by the substitution of an address to the Trinity for an older address to the Son. In this case the phrase Thou didst save mankind by Thy compassion’ finds a natural explanation.

(d-e) Thanksgivings for Incarnation and Redemption. There is nothing of much importance to be said about these clauses, except to draw attention to the parallel with Hippolytus (c) and (d) of the memorials of the incarnation and redemption in Addai and Mari (d) and (e). There is also some similarity of language between Addai and Mari (e) and Hippolytus (e) but the real parallel with Hippolytus (e) in thought is in Addai and Mari (i).

(f) The Presence. This is the first important structural difference of Addai and Mari from Hippolytus. Part of what is put after the institution narrative in Hippolytus (7) (‘because Thou hast made us worthy to stand before Thee’) Addai and Mari places before its own equivalent to an institution narrative. We have already noted the implication of the allusion to Matt, xviii:20, ‘Where two or three are gathered together in My Name, there am I in the midst of them.’ In the reference to ‘standing before Thee’ in Addai and Mari there is probably an allusion to Luke xxi:36 — ‘pray ... that ye may be worthy ... to stand before the Son of Man.’ Behind all this section (f) of Addai and Mari lies the New Testament idea of the eucharist as an anticipation of the second coming and last judgement. (In scriptural language to ‘stand before’ God has often the sense of ‘to appear for judgement.’) But it is all put by way of allusions which are unfamiliar to us, though doubtless conveying their meaning with sufficient clearness to those who used and framed the prayer.

(g) The Institution. Addai and Mari has no explicit institution narrative, but it has an equivalent to it in this brief allusion to what happened at the last supper. The important point to notice is that structurally it plays precisely that pivotal part in the whole prayer which the extended narrative plays in other prayers. It states the authority for performing the eucharist and justifies the petition for communion which is about to follow. The difference of treatment from Hippolytus and Sarapion should not be allowed to obscure this fundamental similarity between the two types of prayer.

(h) Statement of the Purpose of the Eucharist (= Hippolytus (h)). This section of Addai and Mari in its present form has in any case been rewritten, since it suddenly refers to our Lord in the third person, instead of addressing Him directly like the rest of the prayer. The whole connection of thought between (g), (h) and (i) is very confused and difficult to follow. Mr. Ratcliff, emphasizing the parallel between ‘example’ in (g) and ‘likeness’ in (h) is disposed to omit the words ‘great and fearful and holy and life-giving and divine’ in (h) as a later expansion, but to retain the rest of (h) as an original part of the prayer. Interpreting ‘the great and marvellous dispensation’ of (j) as ‘the passion and death and burial and resurrection’ mentioned in (h), he would exclude (i) altogether from the original form of the prayer. He regards its interpolation — at all events in this position — as a later insertion made to bring Addai and Mari more into line with Greek Syrian liturgies (cf. S. James, j1, j2, p. 191).

I confess that I cannot, as at present advised, quite accept this reconstruction, for a variety of reasons. First, this does not help us as regards the sudden ‘switch’ in the address of the prayer from the Son to the Father, about which Mr. Ratcliff offers no suggestion; nor does it mend the halting construction of the whole sentence. It is impossible to be dogmatic in such a case, but it seems to me that the real interruption to the sequence of thought in the prayer lies precisely in this clause (h), with its sudden wordiness and change of address, and its equally abrupt mention of the specific events of ‘the passion, death, burial and resurrection’ which the prayer has carefully avoided mentioning everywhere else. (Cf. e.) The prayer as a whole is concerned with the eternal effects of redemption mediated by the eucharist, not with the historical process of the achievement of redemption in time. If (h) be omitted, the grammar, sequence and intention of the prayer become clearer. The ‘example which is from Thee’ (g) then justifies the petition for communion in (1); the allusion to the last supper (g) explains ‘the oblation’ of the church in (1). As we shall see, there is a close connection of thought between (g) and (i) which would make them complementary in any form of the prayer. I conclude, therefore, despite the acknowledged authority of Mr. Ratcliff on the history of the Syrian liturgy, that it is (h) which is an interpolation inserted to bring Addai and Mari more closely into line with Greek Syrian liturgies; and that (1) is an integral part of the prayer in anything like its present form. Some indication of the importance of the point is that with the elimination of (h) there disappears the only direct reference in the whole prayer to the passion and resurrection of our Lord.

(i) Prayer for Communion. The interpretation of this section is technically a somewhat delicate matter. It is natural that those scholars who accept the theory that some petition that God would ‘send’ the Third Person of the Holy Trinity to ‘make’ the elements the Body and Blood of Christ was an essential part of every primitive eucharistic prayer, should be disposed to see here only one more example of what they conceive to have been the universal primitive practice. It is equally natural that those scholars who believe such an epiklesis-petixion to have been a Greek invention of the fourth century should be inclined to treat the whole section as a later interpolation intended to bring Addai and Mari into line with Greek fourth century developments.

Both ways of regarding it seem rather too simple to fit all the facts of the case. On the one hand, (1) is hardly an epiklesis at all, in that it does not actually pray for any sort of conversion of the elements, but for something quite different, namely for the benefits of communion. It is in fact a petition for those benefits exactly parallel to the clauses we have already found forming the essential petition of the eucharistic prayer before the doxology in Hippolytus (k) and in Sarapion (e2). On the other hand, the terms in which Addai and Mari frames this petition are so obviously primitive (and, I would add, so obviously un-Greek), resting as they do upon that Jewish eschatological doctrine which tended to be lost to sight in gentile Christianity after the second century, that one must hesitate a good deal to regard (i) as any sort of late invention. As regards its later transference from somewhere else in the rite to this point, this is a possibility. But we cannot eliminate this section without cutting out of the prayer as a whole every element of petition whatsoever, which is in itself an improbable form for such a prayer to take after the second-third century.

Finally, while I agree that there is no vestige of evidence in any Greek or Latin author outside Syria during the first three centuries that the Holy Ghost was recognised as playing any part whatever in the consecrating of the eucharist (which in that period is invariably ascribed to the Son), there is one Syrian piece of evidence (cf. p. 278) that ‘Holy Spirit,’ in some sense, was recognised as playing some part in the consecration by Syrian churchmen during the third century, if not earlier. Addai and Mari is not a Greek or Latin document but a Syriac one, and it is best considered in relation to its own special background of Semitic Syrian thought and altogether apart from the ideas of the Greek and Latin churches. We can therefore leave the whole controversy about the Greek epiklesis on one side for the moment, and consider this clause of Addai and Mari simply in what it says itself — ‘May there come, O my Lord, Thy Holy Spirit and rest upon this oblation ... and bless and hallow it that it be to us... for the pardon of oifences ... and for the great hope of resurrection from the dead and for new life in the kingdom of heaven ...’ What exactly is the meaning of ‘Thy Holy Spirit’ here, in a prayer addressed to the Son?

A quotation from the standard work on Jewish theological doctrine, which is remote from all suspicion of partisanship on questions of Christian liturgy, will give us the clue. ‘Christians speak of God’s being in their churches, and of the presence of the Holy Spirit in their religious assemblies or with the individual in secret prayer, without meaning anything different. In Jewish literature also the "Holy Spirit" frequently occurs in connections in which "the Presence" (shekinah) is elsewhere employed without any apparent difference of usage . . . There are certain limitations to be observed in this Jewish equating of ‘the presence’ of God with ‘the spirit’ of God. But it is clear that in the Old Testament ‘the spirit of the Lord’ which brings superhuman strength, wisdom, insight, etc., is not intended to represent a personal agent, but a force — in the older stories often almost a physical force. In general ‘the spirit of the Lord’ is rather a manner of conceiving of God Himself as active in a thing or person, than even a divine attribute. The spirit of the Lord’ seems to refer particularly to God’s presence as energising (and is therefore especially connected with the excitement of prophesying); while the much rarer term ‘the holy spirit,’ though equally impersonal, seems to refer to God’s presence as ‘brooding’ or ‘resting’ on a thing or person, like ‘the cloud’ of the shekinah resting upon the Mercy Seat. Thus in a well-known verse of the fifty-first Psalm, ‘Cast me not away from Thy presence’ is equated with ‘Take not Thy holy spirit from me.’ In the Mishnah there is a tale of a gathering of rabbis at Jamnia, at which a mysterious voice was heard saying, ‘There is here a man who is worthy that the holy spirit should rest upon him, but that his generation is not worthy.’ The Talmud in telling the same story substitutes ‘the presence’ (shekinah) for ‘the holy spirit,’ apparently with no consciousness that it is making any change. Cases are even known in which different MSS. of the same Jewish work use the terms shekinah (presence) and ruh-hakodesh (holy spirit) indifferently in copying the same sentence.

Nor was this conception of ‘holy spirit’ as virtually meaning the ‘presence of God with power’ confined to Judaism. Without entering here into obvious cases of its appearance in early Christian writers, it is enough to point out that it was taken up into the usage of the jews who wrote the Christian New Testament. Thus S. Paul can say of the risen and glorified Lord in heaven now ‘energising’ on earth through His members, ‘The Lord is that Spirit’ (2 Cor. 3:17). And a modern New Testament scholar can sum up a discussion of the Pauline doctrine of the Mystical Body with the words: ‘The Spirit is the element or power whereby the glorified Body or Person of Jesus is present to us and inflows upon us.’

If we may take it that in the very archaic prayer of Addai and Mari the words ‘Thy holy spirit’ applied to the Son are to be understood as the virtual equivalent of ‘Thy presence’ or ‘the power whereby Thy glorified Body is present to us,’ in the fashion of the Old and New Testament writers, the whole construction and meaning of the petition become perfectly clear and straightforward. The prayer is addressed to the Son, Who is reminded of His own ‘example’ given at the last supper. ‘May Thy glorified Body or Person come upon this oblation of Thy servants to bless and hallow it that it may be to us the means of sharing here and now in Thy glorified life.’ Such at least seems to be the only reasonable interpretation of the actual things for which the petition as it stands makes request. I venture to think that this is not a ‘later’ but a very early conception indeed of the results of receiving holy communion, exactly in line with that conception of the whole eucharist as an anticipation of the second coming of our Lord which began to die out in most churches before the end of the third century, or even earlier.

Two small points remain to be noted. First, it may be asked why a petition for the ‘coming’ of our Lord — the Word — in (e.g.) Sarapion should be a later development of the prayer, while in Addai and Mari it seems to be an integral part of the structure. Development varied from church to church, but I think we can see one reason in this case in the different form of reference to the last supper in the two prayers. In Sarapion, as in Hippolytus, the quotation of our Lord’s words of institution sufficed to identify the church’s bread and wine with the Body and Blood of our Lord’s promise, by their actual recitation — This bread is the likeness of the Body because the Lord Jesus took bread saying. .. This is My Body ...,’ as Sarapion puts it. But where, as in Addai and Mari, the reference to what took place at the last supper was in the form of a mere allusion, there was needed further verbal expression of the identification of the church’s offering with what our Lord Himself had pronounced it to be. This is expressed by Addai and Mari in its usual allusive style by the prayer addressed to the Son, ‘May there come, O my Lord, Thy presence upon this oblation of Thy servants.’ Some such petition would be felt to be necessary in eucharistic prayers upon this particular Syrian model from a very early date, in a way not so pressingly felt where an institution narrative could be understood to supply the identification.

Secondly, all that Hippolytus expresses about the nature of the eucharist by calling it the ‘priestly ministry’ of the church, and Sarapion expresses by calling it a reconciling ‘sacrifice’ and by ‘offering the likeness’ of the Body and Blood, is expressed in Addai and Mari by the one word, ‘this oblation of Thy servants,’ which from the context is clearly the bread and the cup. For all its great differences of form and arrangement Addai and Mari witnesses quite sufficiently to the one universal interpretation of the eucharist as sacrifice, even though the hellenistic liturgies have developed this idea more explicitly, as Addai and Mari in turn develops other aspects (e.g. the second coming) which these leave in the background.

(j) The Doxology. Here again an attempt has been made to redirect the prayer to the Father, by the insertion of the words ‘of Thy Christ.’ But we have already learned from (f) that ‘Thy Name’ in which the communicants are ‘gathered,’ and which in (j) is ‘glorified’; is the Name of Jesus, so that the interpolation is obvious. The doxology here is not an ascription of praise to the Three Persons of the Trinity — nothing so theological! It is simply a ‘glorifying of the Name’ in the old Jewish fashion, and a remarkably beautiful one. We may compare it with the very ancient (possibly pre-christian) Jewish prayer known as ‘Half-Kaddish? which in the synagogue ritual marks off the close of separate parts of the service: ‘Magnified and hallowed be His great Name in the world which He created according to His will. May He establish His Kingdom in your lifetime and in your days, and in the lifetime of all the house of Israel speedily and in a near time. May His great Name be blessed for ever and to all eternity.’ In Addai and Mari the world has been ‘re-created’ by the precious Blood, and the Kingdom has been established; the communicants are within it even in this world and they already bless and magnify ‘the living and life-giving Name’ of Jesus for evermore in ‘new life in the kingdom of heaven with’ all the saints, for ‘the great and marvellous dispensation’ of redemption. The eucharist itself is here the direct fulfillment of the old Jewish eschatological hope.

Addai and Mari is obviously peculiar among eucharistic prayers, both in its subtle allusiveness to so much in the New Testament background of the eucharist which other early prayers leave undeveloped, and in its strange ignoring of elements which they explicitly state. To come upon a eucharistic prayer which from beginning to end in its original form has no mention of God the Father or of the Holy Trinity, of the passion of our Saviour or His resurrection, which does not so much as use the words ‘bread’ and ‘wine’ or ‘cup,’ or ‘Body’ and ‘Blood,’ or speak the Name of ‘Jesus’ is in itself remarkable. No less unusual is the omission of any explicit mention of ‘partaking’ or ‘communion.’ All these things are no doubt latent there and taken for granted; but they are not of the framework of this prayer, as they are of the framework of prayers that have been inspired by the systematic Greek theological tradition. Addai and Mari is a eucharistic prayer which is concentrated solely upon the experience of the eucharist, to the momentary ignoring of all other elements in Christian belief and thought. Maranathal ‘Our Lord, come!’ (or perhaps ‘has come’), the ecstatic cry of the first pre-Pauline aramaic-speaking disciples, is the summary of what it has to say.

These things need to be taken into account in estimating the age of this prayer, for the substance of which the later second or early third century hardly seems too early a date. However that may be, it is obviously archaic enough in form and feeling to be comparable with the prayer of Hippolytus from the opposite end of the Christian world and the opposite pole of Christian thought. It is not only in their contents that the two prayers form a contrast, so that what each develops and insists upon the other leaves unsaid or barely hinted at. It is in their whole background of thought and genius that they are different. Hippolytus, for all the relics of old Jewish form, is thoroughly hellenic in its attempt to frame its statement of the essential meaning of the eucharist in rational relation to the whole Christian revelation. Addai and Mari is equally Semitic in the intensity of its absorption in the eucharistic experience, and in its concentration upon eschatology to the exclusion of philosophizing.

But when one has recognized the great differences not only of structure but of mentality which lie behind them, and which demonstrate their wholly independent history, the underlying agreements are the more striking. One need only refer back to the three points we noted as distinctive of the substance of Hippolytus’ prayer to see at once that they are found, perhaps with a different emphasis, but unmistakably the same points, in this wholly different Semitic tradition. (1) The institution at the last supper is central in the construction of the prayer, as the authority for what the church does in the eucharist. The difference in the fullness of reference between the two prayers does not in the least affect the pivotal nature of the reference in both cases. (2) The essence of the eucharist — what the church does in the eucharist — is the oblation of the bread and the cup. This is identified with the Lord’s Body and Blood by His own promise and command, to which Addai and Mari makes a bare but sufficient allusion in the reference to ‘the example which is from Thee.’ (3) The whole rite ‘recalls’ before our Lord, not the last supper, but the redemption He has wrought for mankind, and makes this present and operative by its effects in the communicants.

In Addai and Mari, by contrast with Hippolytus, the emphasis is not on the historical process of redemption by the passion and resurrection, but on its eternal results. That is ultimately the great difference of idea between them; and even this idea, which is emphasized in Addai and Mari, is found in a subordinate position in Hippolytus (e).

We have already spoken of the history of this rite, of which the present text both in Greek and Syriac descends from an Antiochene (? early fifth century) edition and expansion of the fourth century rite of Jerusalem. This older Jerusalem form is known to us only from the account of it given by S. Cyril of Jerusalem to the newly confirmed, who had just attended it for the first time, in Easter week a.d. 348. The Greek S. James will be cited as Jg and the Syriac as Js, and the summary by S. Cyril as C. In the original the passages of C which we reproduce here are absolutely continuous (Catechesis, xxiii- 5-11), though they have to be broken up here in order to relate them to the text of Jg and Js, which has been expanded after S. Cyril’s time. Jg and Js have been revised independently of each other, now one, now the other representing a better text. I follow as a rule Jg, for convenience, noting only some of the variants of Js. Words between + • • • + are not in Js. Matter underlined in Jg is derived from C.

After our discussion of the contents of the prayers previously considered there is no need to comment closely on S. James. The reader will be able to see for himself just how fully and yet how independently (g) (h) and (i) in S. James once more illustrate those three points which we originally noted from the prayer of Hippolytus as containing the essential statement of the meaning of the whole eucharistic action.

But this is in S. James as it is given here, which is substantially a fifth century edition. There are obviously problems concerning the relation of this to (i) the summary of the rite of Jerusalem given by S. Cyril in his Catecheses c. a.d. 350, and (2) the old fourth century rite of Antioch. A full discussion of these problems would involve entering into technical questions of the greatest interest to a specialist but not essential to the purposes of the general reader, and involving many complications. It seems better therefore only to point out quite cursorily some indications of the history underlying the present text of S. James.

S. Cyril’s summary of the eucharistic prayer opens with a preface of which the greater part is recognizable in S. James (a) taken over verbally into its text. There is a curious detail, however, in Cyril’s phrasing which is not taken over by S. James, but which suggests that the Jerusalem preface was originally borrowed from the Egyptian tradition of Alexandria (where the use of the preface and sanctus was probably first developed). The third century Alexandrian writer Origen in treating of the two seraphim in Isaiah vi., in close connection with the eucharistic preface and sanctus, makes it clear that he interprets Isaiah 6:2 as meaning that the two seraphim ‘had each six wings; with twain he covered the Face of God and with twain he covered the Feet of God and with twain the seraph (itself) did fly.’ Accordingly we find the seraphim in Sarapion’s preface (Sar. b1), ‘With two wings covering the Face’ (to prosopon), i.e. of God. By the time of S. Athanasius the Alexandrian church had altered this to the usual later form, ‘their faces’ (ta prosopa), as we find in the text of S. James, and as is attested at Antioch in the later fourth century by S. Chrysostom. But Cyril of Jerusalem, like Sarapion, still keeps to the third century Egyptian interpretation, a sign of the quarter from which the Jerusalem rite had originally borrowed the use of the preface and sanctus.

After the sanctus comes the great puzzle in Cyril’s account of his eucharistic prayer. ‘Next (eita), having sanctified ourselves with these spiritual hymns (i.e. the sanctus), we entreat God to send forth the Holy Spirit.. .’ Is it really possible that in the Jerusalem rite the invocation of the Spirit followed immediately after the sanctus, with no thanksgiving for creation, incarnation and passion, no narrative of the institution or anamnesis clause, or anything else, between? That is what he appears to say, but the statement has appeared so improbable to successive commentators and liturgists that they have all tried hard to make him say something else. So, e.g., Brightman: Cyril ‘is only expounding the salient points of the rite, and for the purposes of his exposition the whole passage between the sanctus and the intercession would be a single paragraph with the form of invocation for its essential point.’ He then goes on to try to find passages elsewhere in Cyril’s writings which ‘may be assumed to represent the contents of the (missing) paragraph.’

I confess I am sceptical of such methods of dealing with a writer who elsewhere shews himself so faithful a summariser. Brightman fails to find a single phrase other than scriptural quotations common to Cyril and that part of the text of S. James which we here label (b-i). One observes, too, that ‘next’ (eita) is one of Cyril’s habitual transitions, and that it invariably means with him what it says — ‘next.’ Thus (23:4-5), after commenting on ‘Lift up your hearts’ and ‘We have them with the Lord,’ Cyril says, ‘Next, the priest says "Let us give thanks unto the Lord" ...’ (and after a comment on this) ... ‘Next, you say "It is meet and right." ‘So in his account of the eucharistic prayer (p. 192), ‘Next, after the completion of the . . . sacrifice, we entreat etc....,’ where the intercessions do actually come ‘next’ in the text of S. James. ‘Next we call to remembrance all them that have fallen asleep,’ where there is good evidence that the clause commemorating the saints did come ‘next’ to the petition "for all that are in need’; and so on. Everywhere else in Catechesis xxiii. when Cyril seems to omit even a few words of the rite from his commentary he appears to insert not ‘next’ [eita) but ‘after this’ (meta tauta) before resuming his summary. I find it difficult to assume that in this one case by ‘next’ Cyril meant ‘After a great part of the prayer has been said.’ And if he did mean that, why associate the invocation so closely with the sanctus: ‘Next, having sanctified ourselves with these spiritual hymns, we call upon God, etc. . .’? He is going through the contents of the prayer for the benefit of those who have just attended the eucharist for the first time in their lives, for whom such skipping about would be quite unnecessarily confusing. On the whole it seems much more likely that Cyril means what he says, and that the invocation in the fourth century Jerusalem rite followed immediately upon the sanctus, however unexpected such an arrangement may be to us, with our modern presuppositions as to the ‘proper’ arrangement of a consecration prayer.

This invocation is of a type we have not hitherto met. There is no room here for the old Syrian equivalence of ‘spirit’ with ‘presence.’ What is intended is unmistakably a prayer for the descent of the Holy Ghost, the Third Person of the Holy Trinity, as at Pentecost. Whether the elaboration on the office of the Holy Ghost now found in (j1) of S. James stood in Cyril’s rite or not, his sixteenth and eighteenth Catecheses make it clear that he held the doctrine of the full Personality and Godhead of the Holy Ghost with a precision and clarity not very common among his contemporaries. (The Godhead and consubstantiality of the Third Person of the Trinity were authoritatively promulgated only in a.d. 381 by the Council of Constantinople, after more than a generation of controversy and confusion on the matter.)

Not only is the invocation itself in Cyril given a precision of address which is lacking in that of Addai and Mari (i), but the petition which follows in Cyril — ‘that He may make the bread the Body of Christ,’ etc. — has been given a different turn to that of the old Syrian invocation in Addai and Mari, ‘that He may bless and hallow it, that it may be to us for the pardon of offences,’ etc., which is really a prayer for the benefits of communion. That of Cyril is a prayer for the means of communion. In Cyril a new idea, that of the ‘transformation’ or ‘conversion’ of the elements, finds clear Liturgical expression.

This is not wholly a revolution. Second century writers like Justin, Irenaeus and Hippolytus could write that ‘the food which has been made eucharist is the Flesh and Blood of that Jesus Who was made Flesh’; the reserved sacrament ‘is the Body of Christ,’ ‘the cup and the bread receive the Word of God and become the Body and Blood of Christ.’ But there is a real step, even if it be an inevitable one at some point or another, from such language to the formulation of a theological theory as to how the identification of bread with Body, wine with Blood comes to be — a theory about ‘the effects of consecration.’ And that step is taken for the first time in the fourth century, and among extant writers for the first time explicitly by S. Cyril of Jerusalem.

It is true that the idea of such a petition is at least half developed in the eucharistic prayer of his older contemporary, Sarapion: ‘O God of truth, let Thy holy Word come upon this bread that the bread may become Body of the Word ...’ The idea of the necessity or desirability of such a petition was ‘in the air,’ as we say, in the first half of the fourth century, perhaps in some circles in the third century. But Sarapion’s language is still linked with older ideas (cf. Irenaeus, ‘The cup and the bread receive the Word of God’). This is, one might say, the product of ‘popular’ rather than ‘scientific’ theological reflection upon the mystery of the eucharist — that the Word Himself, the Second Person of the Blessed Trinity, Whom the communicant receives in communion, should be invoked to ‘come upon’ the elements (in some sense), as He took to Himself the Body formed in the womb of Mary. But Cyril gives clear-cut expression in his liturgy to a different theological theory, which is more evidently a product of the schools: ‘to send forth the Holy Spirit that He may make the bread the Body of Christ... for whatsoever comes in contact with the Holy Spirit is hallowed and transformed.’ After that the way is clear, on the one hand for the development of the idea of a ‘moment of consecration,’ and for the Eastern identification of that ‘moment’ with the invocation — in Cyril’s rite no other possibility could suggest itself — and on the other for a clearer definition of doctrines of ‘conversion’ or ‘transformation’ of the elements, issuing ultimately, by a process of selection, in a particular metaphysical explanation — transubstantiation.

After the invocation Cyril’s rite appears to ‘complete the sacrifice’ (in his own phrase) by an act of offering, as found in the text of S. James (l). It then proceeds to the intercessions, on the ground that ‘this is the greatest aid to their souls, for whom the entreaty is made in the presence of the holy and most dread sacrifice? Once more here is a novelty, or rather two novelties. The idea of the special efficacy of prayer in the presence of the Blessed Sacrament (developed long afterwards in the Teutonic countries of the West in such practices as ‘Exposition’) is here revealed as an originally Eastern notion. So far as I know nothing similar had been said by any author before Cyril. From at least the later second century it had been customary everywhere to offer the sacrifice for particular objects, but the matter had not been further defined. Once again there is not exactly anything wholly revolutionary in what Cyril says, but again there is a logical and (to my mind) a theological step in the process of developing an accepted practice into a theological theory. And again Cyril is the first whom we know to have taken that step.

The other novelty lies in the use of the word ‘most dread’ (phrikodestatos = literally, ‘what makes one’s hair stand on end’) of the consecrated sacrament. This ‘language of fear,’ which Cyril uses in one or two other places, is unexampled in any previous writer treating of the eucharist. Scrupulous care against accidents to the sacrament had been insisted on by earlier writers; they emphasise on occasion that we should ‘fear’ to make an unworthy communion. But they suggest nothing corresponding to ‘fear’ or ‘dread’ of the consecrated sacrament as such. This idea of the ‘awfulness’ of the sacrament, however, soon became a commonplace with Syrian writers (notably Chrysostom) from whom it passed into the Eastern liturgies, though it never took much hold in the West. Again Cyril stands out as the representative of an innovation destined to a long future, not wholly out of connection with the past, but distinctly something new. When we add that Cyril is the first writer to mention the commemoration of saints in the eucharistic prayer (and he has a theological theory about that, too) we begin to understand the sort of man and the sort of rite in the sort of church we are dealing with. The church of Jerusalem in the fourth century is Very advanced’ and S. Cyril is ‘a very extreme man,’ with no overwhelming reverence for old-fashioned churchmanship.

Is such a prayer as his summary seems to describe — preface and sanctus, followed at once by a consecratory invocation, offering, intercessions and Lord’s prayer — a possibility? Or must we believe with the older liturgists that Cyril’s summary omits without trace half the contents of his eucharistic prayer? The reader has the whole of the textual evidence before him. For my own part I believe that he means what he says and has adequately described the whole of his rite.

If so, can we see how such a rite, of so unexpected a form, could come into existence? What has happened to the old ‘thanksgiving’ section which opened the traditional form of the prayer in other churches?

We have already seen that the introduction of the preface and sanctus from Alexandria had in effect destroyed the ‘thanksgiving’ opening in Sarapion’s prayer at Thmuis. The introduction of the preface and sanctus has done the same thing in the present Roman canon. Sarapion’s prayer has filled up its place with its theological hymn (a1 and a2) and its prayer about ‘the living sacrifice’ (c). It seems entirely possible that the introduction of the Alexandrian preface and sanctus at Jerusalem should have had the same sort of result as at Thmuis, but that there the gap was not filled up at all, as it was not filled up at Rome.

But, it may be said, at Thmuis and at Rome the disuse of the ‘thanksgiving’ section still left intact the institution-narrative and what followed. Why are these missing, along with the ‘thanksgiving,’ at Jerusalem? There was in any case no stereotyped line of development in the different churches in the course of such changes; but a particular answer suggests itself in this case. At Rome and Thmuis the reference to the last supper formed a considerable part of the prayer — a narrative. In Syria, if Addai and Mari be any guide, it was a mere allusion to the last supper, which, however pivotal in the structure of the prayer, was from the first supplemented with some sort of petition. Such an allusion could be dropped more easily than a full narrative in the course of an extensive alteration of the traditional prayer, provided that the petition to which it pointed was retained and elaborated in such a way so as to include somehow the allusion to the last supper.

This seems to be roughly what has happened at Jerusalem. If we look back at Addai and Mari for a moment (p. 179), after the allusion to the last supper as ‘the example,’ there comes the petition (1) for ‘holy spirit’ (i.e. ‘presence’) with the ‘offering’ of the elements (in the phrase ‘this oblation of thy servants’). This issues into the petition ‘to bless and hallow it,’ developing into a prayer for the benefits of communion (‘that it may be to us for the pardon of offences,’ etc.). If we look at Cyril’s rite now, it seems that the invocation has been rephrased so as to include the force of both the reference to the last supper and the vague invocation of ‘holy spirit’ on ‘the oblation.’ The change of the petition from ‘bless and hallow it that it may be to us for the forgiveness of sins and eternal life,’ to the exact theological notion ‘that the Holy Ghost may make the bread the Body of Christ’ etc. does recall the last supper by its terms (bread, Body, wine, Blood) in a way that the petition in Addai and Mari (f) fails to do. The offering of the sacrifice in the brief phrase of Addai and Mari, ‘this oblation,’ has been made more explicit in Cyril; and the prayer for the communicants has become Cyril’s unprecedentedly developed intercessions.

I feel bound to point out that the last three paragraphs are in themselves mere speculation, as no other page in this book is speculative. Yet I think it may be claimed that these are ‘scientific’ speculations about facts, in the sense that though we are not able to make a connection between ascertained earlier facts about the third century rite of Jerusalem (of which nothing is known) and the account of it given by S. Cyril, we have to relate Cyril’s rite, unusual as it appears at first sight, quite closely to the general Syrian liturgical background. If his terminology be closely examined, it will be recognized, I think, by anyone methodically acquainted with the development of such things, that it is unmistakably post-Nicene in its key-words. This means that it is in large part a product of some revision not more than twenty years before Cyril commented upon it for the catechumens in a.d. 348. Though each separate item has been equipped with a basis of an up-to-date theological theory, which has largely dictated the actual form of each item in the revised prayer, it would not be quite fair to describe the fourth century rite of Jerusalem as a mere collection of the latest ideas from all over the place, put together into a liturgy without any regard whatever for local tradition. Things did not happen quite in that way in the church before the sixteenth century. For all its superficially novel form, the Jerusalem liturgy is still integrally related to earlier Syrian tradition as this is exemplified by Addai and Mari. (In saying this I do not mean to suggest that Addai and Mari as such was in use at Jerusalem in the third century, but merely that something on the same lines may be taken as by far the most probable form of the earlier Jerusalem use.) In Cyril the old Semitic eschatological tradition of the Syrian eucharistic prayer has been hellenised and ‘theologized’ and transformed, with an obvious desire to be up-to-date and correct. But it is still fundamentally Syrian even in the form in which he describes it. The great influence which the rite of Jerusalem was destined to exert directly and indirectly on all the Eastern rites (and even on some Western ones) during and after the fourth century renders this a fact of outstanding importance.

How far does Cyril’s rite still conform to those basic ideas which so far we have found reproduced so faithfully but in such various ways by the prayers we have studied? There is one difference which stands out — the prayer has been given an entirely new pivot instead of any reference to the last supper — the invocation. But even here the elaboration of its terms to include the words ‘bread,’ ‘Body,’ ‘wine,’ ‘Blood,’ does something to restore the loss. Yet this seemed to other Eastern churches which adopted the Jerusalem form of invocation insufficient to satisfy the traditional sense of the necessity of some clearer allusion to the last supper. We shall find in a moment S. James supplying an institution-narrative from another source, and this is typical of all the Eastern rites which adopted this peculiar Jerusalem form of invocation. In Cyril’s rite there was no option but to regard the invocation as the ‘moment of consecration,’ an idea which was coming in during the fourth century in the East. Elsewhere, by retaining the old institution-narrative or allusion alongside the newly adopted ‘consecratory invocation,’ the Eastern rites laid the foundation of that liturgical and theological duality (not to say confusion) in their theory of the consecration and the eucharistic prayer, which all the efforts of their theologians from Chrysostom to Cabasilas and Mark of Ephesus have never quite succeeded in explaining, or explaining away. It has its roots not in theological theory but in liturgical history.

As regards the other two points, the eucharist is still explicitly something ‘offered’ to God, though it is no longer stated to be ‘the bread and the cup’ which the church offers, but ‘this fearful and unbloody sacrifice.’ It is not easy to say whether the rite is regarded more particularly as the representation of the last supper or of Calvary, because all explicit mention of either event is lacking throughout the whole prayer — a survival of the same sort of Syrian ‘allusiveness’ as we have found in Addai and Mari. If the terms of the invocation recall the last supper, the phrase at the end of the intercessions, ‘we offer Christ immolated for our sins, propitiating God . . .,’ recalls the sacrifice of the Cross. But there is nothing here corresponding to the explicitness of the anamnesis of Christ’s death and resurrection in the prayer of Hippolytus, or of the ‘likeness of His death’ in Sarapion.

But the most important difference between the Roman and Egyptian prayers and those of Syria lies in the absence from the latter of all mention of ‘partaking,’ of actually receiving holy communion. Addai and Mari shares this omission with Cyril, but at least in Addai and Mari there is a prayer for the benefits of communion in its invocation petition (i). Even this has gone from the Jerusalem rite, in the elaboration of its invocation to include the reference to the last supper. No doubt the idea of receiving communion is there in the background, and the practice is presupposed for all present at the liturgy, as Cyril himself makes clear. But this does not alter the fact that the idea of communicating has been ousted from all explicit mention in the eucharistic prayer by the one-sided emphasis on the offering of the sacrifice for various objects, whereby ‘we offer Christ immolated for our sins, propitiating God for them as well as for ourselves’ (xxiii. 7). This is the key-phrase of Cyril’s commentary. A Western massing priest a thousand years later might have been more familiar with this terminology of the fourth century Eastern father than were his own third century predecessors. Again there is here something which one cannot exactly call a revolution. One can parallel both halves of this statement in substance — separately — in the third and even in the second century. But once more Cyril has taken a logical and probably a theological step in advance, not only in combining them, but in framing his exposition of the eucharistic action exclusively in terms of this thought-out theological theory of sacrifice, with no adequate mention of the theology of communion. One can see where things are going along this line — straight to the non-communicant eucharistic piety of the Byzantines and of the later middle ages in the Western church.

To sum up S. Cyril’s liturgy, its ideas are still connected with those of the pre-Nicene past in more than one way, but they are no longer identical with them. They are, however, quite representative of new developments which would carry very great weight in the later fourth and fifth centuries, the period which was decisive in the formulation of later liturgical tradition.

This must be very summarily treated here because a thorough discussion would involve complicated textual questions concerning the relation of S. James to the liturgy of S. Basil, which is not in question in this chapter. It would also require detailed textual comparisons with certain passages in the Antiochene writings of S. John Chrysostom (c. a.d. 370-397) and other evidence. But a number of points can be briefly indicated.

S. James (a). In this preface section of S. James everything seems to be satisfactorily accounted for by the text of the Jerusalem preface in Cyril until we reach the words ‘with ceaseless voices and unsilenced praisings the hymn of victory’ which are not represented in Cyril. It is at least worth noting that these particular phrases are cited from the liturgical preface at Antioch by S. John Chrysostom before S.James had been adopted there.

(b-c). These sections are not cast quite in the form of a ‘thanksgiving,’ but rather of a brief review of sacred history. It would be difficult to give the ‘thanksgiving’ form directly to a narrative which included the Fall. But a mention of Eden and the Fall and the O.T. dispensation generally in this part of the prayer appears to be an Antiochene peculiarity; it is found only in liturgies which derive from the Antiochene tradition. It is again worthy of notice that a similar mention of Eden and the Fall and the Law and the Prophets in this part of the eucharistic prayer is found in Chrysostom’s Antiochene writings.

There is a relationship between S. James (b-c) and the equivalent parts of the liturgy of S. Basil, which is not close enough to describe as ‘borrowing’ on either side but which is nevertheless unmistakable in places. It might well be accounted for by their being independent versions of the same original tradition.

S. James (f, g, h). But this relation is different when we come to the institution-narrative and anamnesis section of S. James. There (after a momentary divergence in/) the texts of S. James and S. Basil are identical, except for the most trifling verbal changes. One rite has directly borrowed off the other, and it appears to be S. James which is dependent on S. Basil. A full institution-narrative was certainly already to be found in the Antiochene rite in the time of Chrysostom, who attributes to it a central importance in the rite. So far as they go, his quotations agree with the present institution-narrative of S. James (g), but this could be due to a common use of 1 Cor. 11 as the basis of the account. There seems to be no trace of an anamnesis section in Chrysostom, and all account of an anamnesis is missing from the verbose description of the rite of Mopsuestia (of Antiochene type) by his contemporary Theodore. If Addai and Mari be an adequate guide, it was precisely the institution-narrative which would need amplifying and the anamnesis section which would have to be supplied from somewhere else in an old Syrian tradition, if this were being brought up to date in accordance with most other Greek liturgies in — say — the fourth or fifth century. This would account for the borrowing here in S.James.

One notices the eschatological emphasis of the latter part of (h) in S. James (cf. Addai and Mari f), including the vivid touch — ‘Spare us O Lord our God’ — which represents the last judgement as actually taking place. Evidently the Syrian tradition which understood the eucharist as an anticipation of the second coming had not weakened when this prayer was composed.

S. James (i) goes on to offer the sacrifice in a single phrase, and then to pray for the forgiveness of sins and ‘Thy heavenly and eternal gifts,’ in substance though not in phrasing very much as in Addai and Mari (i).

It seems worthy of attention that if a doxology were appended after the words ‘them that love Thee,’ we should have in S. James (b-i) a complete eucharistic prayer, parallel in content to but verbally independent of the eucharistic prayer of Hippolytus. Such a prayer would also have a good many points in common with Addai and Mari. But here there would also be the big differences that S. James (b-i) contains a complete institution-narrative and an anamnesis (probably derived bodily from S. Basil) but no invocation of ‘holy spirit’ in any form (up to this point). None of this matter (b-i) is derived from Cyril’s Jerusalem rite, but some of it has distinct points of contact with the scattered allusions to the fourth century rite of Antioch in Chrysostom.

"S. James (j, h). However, S. James in its present form goes on to add an invocation — in fact, as we have seen, two. One of these (j2) evidently contains matter derived from the Jerusalem rite described by Cyril. The other (j1) is in a form which there is some reason to believe was in use in the region of Antioch in the later fourth century, since it reappears in substance in the invocation of the liturgy in Ap. Const., viii. It is also clear from Chrysostom that an invocation of some kind was already in use at Antioch in his day, though it seems impossible to make out the text from his allusions. But one notes that both invocations in S. James come after the point at which the analogy of other rites would lead us to expect such an invocation to be placed (i.e. one would expect an invocation in S. James (i)9 following the words ‘beseeching Thee’ in its first sentence).

"S. James (k). In (k) S. James produces a second prayer for the communicants in the same terms, ‘for the forgiveness of sins and for eternal life,’ as that in S. James (i). With S. James (k) we may compare the prayer for the benefits of communion in Addai and Mari (i). But the brief allusion in the latter to Thy church’ has been expanded in S. James (k) into a rudimentary intercession for Thy holy catholic and apostolic church.’ There is evidently a good deal of duplication in all this part of the rite; there are two invocations, two prayers for the benefits of communion, two offerings of the sacrifice, two prayers for the ‘holy catholic and apostolic church,’ and soon.

S. James (l, m, n) are mostly taken over from the fourth century Jerusalem rite.

One general inference which seems to impose itself from this brief survey is that the fourth century Jerusalem rite was fused with the fourth century rite of Antioch to produce the patriarchal’ rite of Antioch (the present S. James) rather by way of addition to the Antiochene local tradition than by way of substitution for it. Considerable fragments of the supposedly ‘lost’ old rite of Antioch are to be found embedded in the present text of S. James,

Their discernment, however, is likely to be a more complicated matter than the mere subtraction of what can be detected as ‘Jerusalem’ material by comparison with Cyril. There seems to have been more than one stage in the process of compilation to form the present text of S. James, and the details of the process can hardly be accurately disentangled in the present state of the materials. In trris connection I would draw particular attention to the place of the’non-Jerusalem’invocation material in (j1) and (j2) (which has attracted to itself the similar material derived from the Jerusalem rite). Instead of coming in (i) where on the analogy of other rites we should expect it, it is placed as a sort of appendix to the body of the remains of the old Antiochene eucharistic prayer, after the point at which one would look for a doxology to the old Antiochene prayer. This is interesting, because Mr. Ratcliff has pointed out that there are traces of a third century Syrian practice of placing an invocation of the Spirit outside the eucharistic prayer proper, immediately before the fraction. If the present order of S. James preserves (as it seems to do) the outline of the old Antiochene rite, this may have been the original position of the invocation when it was first introduced at Antioch. Strange as it may seem to us with our presuppositions, such a position is really not an unnatural one. The Nestorians of Malabar in the later middle ages inserted the institution-narrative, which their own rite (Addai and Mari) did not contain at all, in that very place just before the communion. They had come to realize that other churches valued and used it and they wanted to include it somehow in their rite, but there seemed no suitable position for its insertion within the structure of their own traditional eucharistic prayer. When many Syrian churches were making such an invocation the central pivot of their rite, Antioch, the mother church of Syria, might well feel that something of the kind ought somehow to find a place in its own rite, and yet be unwilling at that time to disturb its own traditional arrangement of the prayer in this particular matter. A ‘supplementary’ position for new items, after the eucharistic prayer proper and before the communion, is a common form of compromise attested in all rites. (The position of the Lord’s prayer is an obvious example.) In course of time such supplements are always apt to be fused into a single whole with the original body of the prayer, or at least to be treated as inseparable from it, by mere invariable association (cf. the position of the Lord’s prayer at Milan, between the conclusion of the eucharistic prayer and its doxology).

Be that as it may, the evidence of duplication and conflation in all this part of the eucharistic prayer of S. James seems undeniable. Whatever the exact explanation, we have here plain traces of the complicated sort of process by which during the fourth-fifth centuries the great historic rites gradually assumed their final form.

The reader has now seen something of the evidence for a great diversity in the local traditions of the eucharistic prayer during a period which may be roughly defined as from about a.d. 200 to 400. Had the last chapter included even a summary analysis of other prayers, such as the Eastern liturgies of S. Basil (from Asia Minor) and S. Mark (from Alexandria) or the Roman canon, all of which contain a good deal of older material overlaid by fifth and sixth century revision, the impression of a great early diversity in eucharistic prayers would have been strengthened, and the range of ideas found in them would have been extended. We have also seen how towards the close of the fourth century, as a result of continual local revisions and mutual borrowings, eucharistic prayers everywhere were beginning to shew a general structural similarity and even a partial identity of phrasing.

It will be one of the most important technical tasks of liturgical studies in the next ten years to pierce this later superficial uniformity and to recover the fragments of genuinely ancient local traditions beneath. But this is a task which is only beginning to be attacked with properly scientific methods, and it would be out of the question to attempt here even a sketch of the problems which will have to be re-examined in detail by experts before we shall have reached the stage of solidly established conclusions. That would require a book in itself, and one of a much more technical character than this can claim.

Yet it seems necessary, even in a book for the general reader and at the present stage of research, to attempt to give some sort of answer to the main question: Can we hope to penetrate through this (fourth-fifth century) period of growing uniformity, and behind that through the period of the unordered growth of local traditions (in the third-fourth century) back to some sort of original uniformity? Can we hope to find in the primitive church, say in the second century, coherent universal principles which can guide our own ideas about liturgy? Was there anything, for instance, in what is vaguely called ‘the early church’ which might serve as a standard or model by which the perplexities of Prayer Book revision in twentieth century England might be lessened? That is the sort of question which the plain churchman or the practical bishop wants to put to the liturgical student, and to which (so it seems to me) he is entitled to expect a plain and practical answer — and to which (so it seems to him) he does not always get one. I hope I shall not seem to be trying to evade the question if I begin by pointing out the conditions in which such a plain and practical answer has to be framed at present, especially by an Anglican.

The early evidence on the eucharist is both fragmentary and complicated. Not only its interpretation but its discovery is often a matter needing a very delicate discernment. The pre-Nicene church was a secret society, which deliberately intended to seclude knowledge of its liturgy from all but its own tested members. It is as a rule only by hints and allusions that liturgical matters are referred to by writers of the first three centuries in works which deal primarily with other aspects of the Christian religion. (There are exceptions, like Hippolytus’ Apostolic Tradition, but these are few.) To those who frequented the Christian rites such allusions were enough to illustrate the author’s meaning; to others they would convey little or nothing — and the modern student is often among the ‘others’ for practical purposes. It is not surprising, though it is unfortunate, that for two centuries experts have interpreted this sort of evidence in different ways, and that different general theories have dictated two different types of answer to this main question which the plain Christian wants to put. The two schools may be distinguished here as the ‘traditionalist’ and the ‘critical.’ Without going at all deeply into the controversy between them, it is necessary to say a little about their respective theories.

Beginning so far as modern times are concerned with the German scholar Probst about 1860, the traditionalists have for nearly three generations now been proclaiming to such of the public as take an interest in these things that a primitive standard type or model of the eucharistic prayer did exist, and that its form is not difficult to reconstruct. The attempt to demonstrate its existence and explain its meaning has preoccupied most of the more ‘popular’ literature (if that adjective is applicable to any of the productions of liturgists) on the subject for at least sixty years past. Some writers of this school have contended that there existed a ‘lost text’ of the eucharistic prayer, of apostolic or sub-apostolic origin, from which all the historic rites were developed by a process of expansion or perversion. The greater part of the traditionalists, however, impressed by the evidence for a general custom of more or less free phrasing of the eucharistic prayer by the celebrant, have sought rather to establish the idea that there was a normal or standard outline or framework of the prayer, to which all such prayers ought to conform, and to which, they argued, the majority of such prayers have conformed since very early times. This authentic model the earlier representatives of this school mostly found to be best represented by the Byzantine or North Syrian type of prayer, whose earliest complete example is the eucharistic prayer of the liturgy in the Apostolic Constitutions, Bk. viii., from the region of Antioch c. a.d. 375. More recently they have concentrated their attention on the eucharistic prayer of Hippolytus, which is now known to have been one of the sources used by the compiler of the Apostolic Constitutions.

This theory is currently associated in England with the name of that very distinguished liturgical scholar the late Bishop Walter Frere, C.R., whose last book, The Anaphora (S.P.C.K., 1938), may be taken as its latest and most brilliant exposition. But the theory is in reality much older than Frere’s rehabilitation of it, and far from being a peculiarly Anglican thesis. It was first put forward in a fully developed form by the French liturgist Pierre Le Brun in his Explication de la Messe in 1726, but in essentials it goes much further back. It is, for instance, the basis of the anti-protestant polemics of the first editor of the Apostolic Constitutions, the Spanish Jesuit Francisco Torres in the sixteenth century. In a naive form it can be traced back into the roots of the middle ages, to the Carolingian liturgists of Gaul in the ninth and tenth centuries.

In modern times it has attracted the support of three outstanding representatives of German scholarship in three successive generations: Probst (Roman Catholic), Paul Drews (Lutheran), and Dr. Anton Baumstark (Roman Catholic), besides a large number of lesser names not only of the German but also of the French and Italian liturgical schools (e.g. Dom Cagin). In Anglican liturgical study this has been the dominant theory at least since the compilation of the second Scottish Prayer Book in 1764. Its influence here may be traced chiefly to the work of Bishop Thomas Rattray, whose essay on The Ancient Liturgy of the Church of Jerusalem was published in the year after his death, 1744. It is sometimes said that this was the theory generally held by the English Caroline divines of the seventeenth century, but this is true only with such qualification as to be virtually untrue. The fact is that the Carolines, like the Non-Jurors after them, took only an unscientific interest in the early history of the liturgy, and did not advance to the stage of producing serious theories about that, though they had plenty to say about its theology.

Whether its influence in England began in the seventeenth or the eighteenth century, the traditionalist theory has long enjoyed here two great practical advantages for its propagation. As the established and dominant theory, it has affected nearly all the elementary manuals and text-books, so that every fresh exposition of it could always appeal to that general background of liturgical knowledge which most of the clergy had picked up in the course of their professional training. And in itself it offers a clear and attractive theory which anyone interested can grasp without much difficulty, and which can be illustrated effectively by much of the evidence from the fourth and fifth centuries.

Over against the traditional school, however, there stands not so much a ‘school’ as a long succession of some of the greatest names in the history of liturgical scholarship — Tommasi in the seventeenth century, Forbes of Burntisland and Ceriani in the nineteenth, Brightman, Armitage Robinson and Lietzmann in the twentieth, and above all, Edmund Bishop (perhaps the greatest of all liturgists) — all of whom have either explicitly rejected the traditional theory as seriously misleading, or at least based their own studies on a quite different understanding of the evidence. Some of them (e.g. Bishop and Ceriani) had hinted at the possibility of a radical dualism in liturgical origins. In our own day Lietzmann has boldly developed this into the idea that there were from the first two quite different types of liturgy in the church, different not only in form but in essential meaning, which he would derive respectively from the Pauline and the judaising churches of the apostolic age. The critical school (if such they can be called) have differed considerably among themselves in their positive statements, but they at least agreed in this, in rejecting both the form and the basis of the traditional theory of a single primitive type of prayer. They all emphasized the signs of a very great variety in the outline of the eucharistic prayer before about A.D. 350.

Unfortunately, excepting Lietzmann, every one of these names is that of a writer who was very much a ‘scholar’s scholar.’ Their most important contributions on this particular subject are mostly, either like those of Tommasi and Forbes, incidental statements found in works on other aspects of liturgy which are now unprocurable even at second-hand, or else printed as articles buried away in back numbers of theological periodicals which are not very commonly available. And just because their criticisms of the accepted theory are based chiefly on the earlier evidence which is particularly difficult and complicated to handle, their work as a rule shows little consideration for the wayfaring man. The scholar’s caution and perception of nuances, his wariness of the over-simplification of complex questions, his distrust of short-cuts to results, are all qualities necessary for the pursuit of truth. But they do not make for easy reading, and these writers suffer from all these virtues. It is possible to detect in them a sense (eminently reasonable in the state of the evidence until just the last few years) that the main questions of eucharistic origins were by no means ripe for positive solution; and they do not as a rule give more than hints of where they believe the true solutions to lie. The only attempt at a general exposition of a ‘critical’ thesis which has ever been made, Lietzmann’s Messe und Herrenmahl (Bonn, 1926), fully justified this caution. It is spoiled, for all its brilliance, by not a few extravagances.

It is not surprising, I think, that confronted on the one hand by a long-established theory which is attractive and lucid in itself, and which can account for an impressive selection of what passes for ‘ancient’ evidence (though it is almost entirely post-Nicene); and on the other hand by what seemed to be a recondite and chiefly negative criticism, the bulk of what might be called ‘interested but not expert’ opinion in Anglican clerical circles should have tended for many years past to accept the traditionalist thesis without much hesitation. Such outright rejection of it as there has been was derived from attachment to present Anglican liturgical practice, or from post-Tridentine doctrinal sympathies among a certain section of ‘Anglo-catholics,’ much more than from reasons of history or technical liturgical study. The results of this state of affairs became obvious and practical in 1927-28.

We are not here concerned at all with the question whether the proposed new Anglican canon drawn up then was or was not desirable in itself, but simply with the fact that it was the product of a particular technical theory about the early history of the liturgy which had been in debate among scholars for two centuries before 1928, and which at the least had been shewn to be open to serious historical criticism. This does not seem to have been clearly understood by the majority of the bishops when they put forward their proposals, and not at all by the church at large when these were being considered. It was soon obvious that the criticisms of this element in them made by scholars of the caliber of Armitage Robinson and Bright-man greatly surprised and disconcerted men like Bishop Headlam of Gloucester, who were lending intelligent support to the proposals, but who on technical questions of liturgy could speak only as amateurs, as was plain from their replies.

Yet the constructive weakness of the critical school of liturgists was illustrated once more in this, that though they made many incidental suggestions for the practical improvement of the proposed rite, they produced no easily understood criticism of its form or general justification for their own ideas, and no alternative scheme as a whole. In the event their criticisms were ignored by authority as ‘unhelpful’ — a verdict which had in it a certain rough-and-ready justice, but little wisdom, as the issue proved.

This same attitude of surprise tinged with resentment was noticeable in these same interested but inexpert circles ten years later, at the very cool reception accorded to Frere’s book on The Anaphora by the reviewers (mostly competent liturgical scholars) almost without exception in the learned periodicals of all countries. It was inevitable from the form in which Frere had cast his book that discussion in England should reawaken some of the polemics about 1928. It was quite unnecessarily unfortunate that camp-followers on both sides tried to involve a matter of pure scholarship in questions of personalities and ecclesiastical politics. But apart from the small groups which acted in this way, there was a large body of thoughtful Anglican opinion which was genuinely puzzled that such a book should be received by scholars as The Anaphora undoubtedly was, with a virtually unanimous rejection of its main thesis, accompanied by respectful compliments on the manner of its presentation.

Frere himself, as his last letter to me shewed, was by no means unprepared for this reception. He was quite aware that with the advance of knowledge and method in the last twenty years the historical difficulties which confront the traditional theory of a single original type of eucharistic prayer had grown more and more formidable, and that he was probably the last living scholar of the first rank to maintain it in anything like its traditional form. The truth is that the book is a skilful rearguard action, an attempt to recast the traditional theory in such a way that it should still be tenable in face of the growing critical difficulties. It is proper to say that, in the judgment of most of those qualified to pass an opinion, his attempt in the particular form in which he made it must be held to have failed; though it was well worth making and in some things has pointed the way to a truer solution. But in view of the way in which the whole matter has sometimes been handled it seems right to insist here that it is only incidentally connected with the name of Bishop Frere or the proposals of 1927-28, and not at all with doctrinal or ecclesiastical allegiance. It is part of a technical debate among liturgical scholars which had been proceeding at intervals for some two centuries before 1928, though in the opinion of most competent scholars it is now in sight of a conclusion. The theory which Frere embraced originated with the Roman Catholics Torres and Le Brun, and has numbered among its modern defenders Roman Catholics, Lutherans and Anglicans just as indifferently as it has numbered them among its critics.

It will have been worth while reflecting a little at length on this episode if it makes clear the difficulty at the present moment of giving ‘plain and practical’ answers about the primitive eucharistic prayer, of the kind which I for one believe that liturgical science ought to be able to give. The traditional theory did give such an answer, but there is good reason to fear that it was a very misleading answer. On the other hand, the critical school, while it has made good its thesis of a great diversity in pre-Nicene eucharistic prayers and overthrown the traditional theory that the Syrian type of eucharistic prayer represents the original universal type, has found nothing very coherent to put in its place as a plain and practical guide for the modern church. Yet to say, as some scholars have implied of late, that we cannot rightly look to the primitive church for such guidance, because it had not itself achieved any intelligible principles in liturgy, would be, I believe, to consent to a mere reaction against the traditional theory which is not warranted by the evidence. And it would rob the science of liturgy not only of all practical value to the church, but of its chief interest in the eyes of all but a few specialists who might continue to make it their hobby.

Yet if the question continues to be put in the way in which the traditional theory has for so long encouraged the ecclesiastical public to put it, ‘Can we find in the primitive church a model or standard for a modern eucharistic prayer?’ — the answer of the liturgists will be, ‘Certainly not, if what we are required to pursue be any form of the mediaeval or modern myth of a single apostolic or sub-apostolic text of the prayer.’ Such a text never existed, and it is hard to see any complete scheme of a common arrangement in the immense variety of the early material, as this is now slowly coming to light. Yet the pre-Nicene church was quite well aware of what it supposed itself to be doing when it celebrated the eucharist. It should be quite possible to discover and interpret its liturgical principles truly, if only we look for the kind of principle which was then recognized, not those which the fourth and fifth century fathers in their very different situation, or the Byzantines and the mediaeval Latin church, or Tudor and Stuart statesmen, successively elaborated for themselves. Whether pre-Nicene liturgical principles, if we can discover them, will be of much use to us in our very different circumstances is a matter which might require further consideration when we find out what they were.

For the liturgical scholar the technical question resolves itself into this: Does that great variety which has been discerned in the eucharistic prayers of the early fourth century, and which seems to increase as we penetrate back into the third, does that go back all the way to a beginning in the apostolic age in a sort of liturgical anarchy? Or is there some element of truth in the discredited traditional theory of an original uniformity, by which we may find general principles which will interpret the apparent confusion of these prayers? This book has been written partly in order to shew that there is.

What was fixed and immutable everywhere in the second century was the outline or Shape of the Liturgy, what was done. What our Lord instituted was not a ‘service,’ something said, but an action, something done — or rather the continuance of a traditional Jewish action, but with a new meaning, to which he attached a consequence. The new meaning was that henceforward this action was to be done ‘for the anamnesis of Me’; the consequence was that ‘This is My Body’ and ‘This cup is the New Covenant in My Blood.’ Apart from these statements, the formulae which Jesus had used at the last supper, the Jewish grace before and after meals, had referred exclusively to the old meaning. Beyond these brief statements, both the new meaning of the action and the words in which to express it were left to the church to find for itself, and there was nothing to suggest that this was a process to be completed by the first Christian generation.

We have seen that the church in reflecting upon this legacy from her Lord was soon led to disencumber this Jewish action from everything in its traditional Jewish setting which could obscure its new Christian meaning, and so to form the rite of the eucharist apart from the supper. The universal scheme of this, that ‘four-action shape’ in which the prayer formed the second item, went back to the end of the first century, perhaps to the last years of the apostolic generation itself. From the uniformity of this outline everywhere and the early identity of the dialogue introducing the prayer, one would infer that the new form of the rite, together with its new name of ‘the eucharist,’ spread all over Christendom in the last quarter of the first century from a single center, which — if we must try to locate it — is most likely to have been Rome.

What would form the chief content of ‘the’ prayer, which originally afforded the only possibility of giving verbal expression to the meaning of the rite as a whole?

First, the name ‘eucharist,’ ‘thanksgiving,’ governed the whole rite from beginning to end. Secondly, this expressed the old meaning with which our Lord Himself had ‘done this’ at the last supper. Thirdly, this was something carried over from the very roots of the eucharist in the chaburah supper into its new Christian shape, by the retention of the dialogue of host and guests (‘Let us give thanks unto the Lord our God’) as well as by the derivation of the eucharistic prayer from the Jewish berakah (= ‘thanksgiving’). Fourthly, this Jewish berakah itself, traditional at the last supper and the primitive Jerusalem eucharist when this was still celebrated as the beginning and end of a meal, contained elements which looked beyond that mere thanksgiving for food which would soon come to seem quite inadequate as the fullness of the new Christian meaning began to be understood.

When we look back at this berakah (p. 53) and place beside it the consensus of the second century evidence as to the contents of the Christian prayer, we can perhaps see a parallel of thought which does not seem to me to be either fanciful or accidental, though others must judge for themselves.

Its first paragraph opens with the usual formula of address to God in such blessings: ‘Blessed be Thou, O Lord our God’ etc. Besides the specific ‘thanksgiving’ for the meal (which would be irrelevant to the ‘four-action shape’ of the eucharist) it contains a ‘blessing’ or ‘glorifying of the Name’ of the kind obligatory in all Jewish blessings.

It is, however, the second paragraph which is of most importance to us now.

It is quite open to anyone to say that the parallels here are both too vague and too subtle to be anything but accidental. Yet if a prayer had been handed down in a tradition by a process of more or less free reproduction extempore Sunday by Sunday for a century through a long line of celebrants, the most that could be expected to maintain itself would be a series of themes in a certain connection. And this particular series of themes, apparently in approximately the same order, is found as matter of the eucharistic prayer at Rome in Justin c. a.d. 155 and in Hippolytus fifty years later. The same themes, in approximately the same order, are found too in other traditions, e.g. at Antioch and Edessa; though we cannot in these other cases prove that they were in use in the second century, as we can at Rome. Such a widespread use suggests a very early diffusion. And some explanation is required for the fact that the allusion to the last supper in most rites is curiously placed, coming out of its historical order, after the thanksgiving for redemption by the passion.

Despite certain difficulties, it does seem that those who believe that there was an original authoritative outline of the prayer could make out (by a comparison of traditions) an overwhelmingly strong case for regarding this series of Thanksgivings’ as the original opening of the prayer (after the preliminary ‘Naming’ of God), especially if its derivation from the second paragraph of the berakah be admitted. The traditional school have tended for some reason to ignore this series of’Thanksgivings.’ But I will venture to prophesy that this will eventually prove to be their fortress, which the critics will be unable to capture.

The connection — if such there be — between the Jewish and Christian thanksgiving is one of ideas and form only, not of phrasing. The berakah has been entirely re-written in terms of the New Covenant. It concentrates in a remarkable way on the work and Person of our Lord, even where, as by Hippolytus, it is addressed to the Father and not to the Son, as in Addai and Mari. The series is, in fact, in itself an anamnesis of Him, as our Lord had ordained.

On the other hand, if this ‘Thanksgiving series’ (following the preliminary ‘Naming’ of God) formed the original opening of the prayer, it was from quite an early date — let us say vaguely the late third or fourth century — not the only form such an opening could take. An opening sequence of ‘Thanksgivings’ does not appear at all in the only extant examples of the old Egyptian tradition, viz., Sarapion, and the authentic text of the liturgy of S. Mark as found in the Strassburg papyrus (fourth-fifth century).

As this text has not hitherto been given, but will now be necessary to the argument, we may say that a collation of this papyrus, where it is legible, with the mediaeval Greek and Coptic texts of S. Mark reveals the following as having been the opening of the Alexandrian prayer in the later fourth century:

1. Address: ‘It is truly meet and right, holy and fitting and expedient for our souls, O Living God, Master, Lord God the Father almighty, to praise Thee, to hymn Thee, to bless (eulogein) Thee, to confess Thee night and day,
2. Creation: ‘Thee, the creator of heaven and all that is therein, the earth and all that is on earth, the seas and rivers and all that is in them; Who didst create man according to Thine own image and likeness. Thou didst make all things by Thy Wisdom, Thy true Light, Thy Son our Lord and Saviour Jesus Christ:
3. (3a) Preface (1st half): Through Whom unto Thee with Him and with the Holy Ghost, we give thanks (eucharistountes) and offer the reasonable sacrifice of this bloodless worship, which all nations offer unto Thee from the rising up of the sun even unto its going down, from the north even unto the south; for great is Thy Name among all nations and in every place incense is offered unto Thy Holy Name, and a pure sacrifice, offering and oblation,

(3b) Preface (2nd half): cFor Thou art far above all principality and power and rule and dominion and every name that is named ...’ [and so through the rest of the preface to the sanctus, almost verbally as in Sarapion b1; cf.p. 163].

There is here no sequence of the ‘thanksgiving’ themes. But it is conceivable that something of the sort once stood as the opening of this Egyptian tradition as well as of all others. S. Mark (2) looks like a survival of the ‘creation theme’ following the preliminary ‘Naming’ of God, even though it is cast rather in the form of a ‘praising’ for creation than a ‘thanksgiving’ for it (cf Sarapion a1 and a2). This latter word does not appear in S. Mark until we reach (3), and not at all in Sarapion till the end of the prayer (1). It looks as though this ‘thanksgiving’ (it is convenient to retain the word, even though it is not quite accurate in the case of S. Mark) for creation, which is rather pointless as it stands, was once followed by others for the incarnation, redemption, etc. on a scheme comparable to that of Hippolytus and Addai and Mari; and as though the later members of the series had been ousted by the preface and sanctus. But it is to be remembered that the preface and sanctus were already found in the Alexandrian rite at some point by the time of Origen c. a.d. 230, and that there is nothing to suggest that their use was then a recent innovation.

It is usual to regard the preface and sanctus as a peculiar development of the ‘thanksgiving series’ opening of the prayer. But the fact remains that it appears in practice not as a development of it but as an alternative to it, a sort of liturgical cuckoo, which ends by taking the place of the ‘thanksgivings’ whenever it is admitted into the prayer. Only in the prayers of the Antiochene type has a successful effort been made to fuse both forms, by prefixing the preface and sanctus (borrowed from Egypt via Jerusalem) to the old Antiochene ‘thanksgiving series’ (cf. S. James, pp. 188 sq.); and even there, if the wording of S. James b, c, d, be examined, it will be found that the prefixing of the sanctus has led to the elimination of the actual ‘thanksgiving’ form of the clauses. The word ‘give thanks’ has been replaced by the form ‘Holy art Thou,’ etc. When the preface and sanctus were adopted by other churches, as at Jerusalem and at Rome, it displaced altogether in their rites that sequence of ‘thanksgivings’ which Addai and Mari and Hippolytus assure us was the pre-Nicene tradition of Syria and Rome alike, but of which Cyril at Jerusalem and the present Roman canon know nothing.

It seems probable when we look at S. Mark that something of the same sort happened in the first instance at Alexandria itself, where, so far as we know, the preface and sanctus originated. But there the first member of the old Alexandrian sequence of ‘thanksgivings,’ that for creation, survived when the following ‘thanksgivings’ for the incarnation, redemption, etc. were eliminated in favour of the preface and sanctus. Perhaps that for creation survived in S. Mark chiefly through the difficulty of disentangling it from the ‘Naming’ of God in § 1. The opening of S. Mark (in §§ 1 and 2 taken together) constitutes a ‘Naming’ of God as Father and Son, to the exclusion of the Holy Ghost, of the type found as the opening of Hippolytus and Sarapion But it would be difficult to extract the creation theme from the text of S. Mark as it stands, while leaving this ‘Naming’ as a coherent sentence. If we are right in supposing that a series of such thanksgivings once came between that for creation in S. Mark 2 and the preface and sanctus in 3, it would seem that the combination of preface and sanctus with the sequence of ‘thanksgivings’ differed at Alexandria from that found at Antioch. At Antioch in S. James the preface and sanctus come first. At Alexandria in S. Mark the preface and sanctus appear to have come after the sequence of ‘thanksgivings.’ I will hazard a suggestion as to why this should be so in a moment.

To revert now to the general question, Was there an original uniform type of eucharistic prayer? We have found something of which traces appear to be present in all the early traditions, viz:. — An opening address and ‘Naming’ of God, followed by a series of’Thanksgivings’ or ‘Praisings’ on a sequence of themes beginning with creation, incarnation and redemption. (We need not at this point try to decide exactly where this sequence ended, and whether it originally included a reference to the last supper or not. The universal existence of such a sequence is sufficient for our immediate purpose.) But it is when we pass beyond the possible contents of this sequence of themes into the second half of the prayer that the difficulties in the way of establishing the existence of any original universal model of the prayer become really formidable.

The evidence we have already surveyed represents the traditions of the three leading pre-Nicene churches of Syria, Egypt and Rome, and includes all the most ancient evidence extant, except that to be derived from certain heretical gnostic writings. When one has eliminated from the second half of each of these prayers all that can safely be ascribed to later local developments and to borrowings, it is not easy to detect any single scheme upon which they all arrange their parts and ideas.

To take but one instance, though a cardinal one: Three ideas which Hippolytus keeps distinct and arranges in three successive statements (fg, h, i) — the recital of the institution, the anamnesis of ‘His death and resurrection’ and the offering of the bread and the cup — Sarapion in Egypt expresses inextricably entangled with one another in his section d (with no mention of the resurrection). Addai and Mari in Syria contains the first and the last, but in its earlier form, apparently, not the second. But it expresses them differently again, by the barest allusions, in connection with other ideas, in g and i. One can trace in the second half of all these prayers the recurrence of some ideas which are the same in substance, but differently handled and differently arranged. The one obvious point of arrangement in which they all agree in their second halves is that all end with a doxology or ‘glorifying of the Name.’

Thus the later traditions of the prayer all show a similarity of arrangement in their first half, the ‘Thanksgivings.’ Especially impressive is the identity of the series of themes everywhere. But they shew great diversities of content and arrangement in their second half. The inference is that any original material common to them all covered only the first half and the concluding doxology.

Is it possible to conceive of a primitive type of eucharistic prayer which consisted simply of a ‘Naming’ of God, followed by a series of ‘Thanksgivings’ for the New Covenant and concluding with a ‘glorifying of the Name’? It would be without much which later ages considered essential to such a prayer. But at all events one can see how it could be called ‘the Thanksgiving." And after studying the themes of the ‘Thanksgivings’ as they are actually handled in the various traditions, one can see how they could be regarded precisely as ‘the anamnesis’ the solemn ‘re-calling’ before God, of the work and Person of Jesus Christ. Finally, for my own part, I can see how such a prayer as a whole could be derived directly from that Jewish berakah which was used at the last supper, and in the Jewish apostolic church. Such an outline of the prayer could very well be a part of that fixed ‘four-action shape’ of the liturgy by which the chaburah ritual was so delicately adapted to the new Christian form, and which took over amongst other things the very dialogue which immediately preceded and introduced the berakah.

This is all quite possible, but a little evidence is worth a great deal of plausible speculation. Can we find any examples of this type of primitive prayer? The two oldest prayers we have, Hippolytus and Addai and Mari, can both be dated in substantially their present form soon after a.d. 200, and these are both prayers which have a fully developed ‘second half. It will therefore be of little use seeking beyond the second century for an unexpanded prayer. Second century evidence is scanty and hard to interpret, but we can only examine once more our three traditions.

Let us look back at the Alexandrian liturgy of S. Mark, with (1) its ‘Naming’ of God; (2) thanksgiving for creation; (3) preface and sanctus. If — it has not been demonstrated and the reader must judge for himself of the probability of the hypothesis — but if in S. Mark a series of similar ‘thanksgivings’ for incarnation, redemption, etc., originally stood between the present thanksgiving for creation (2) and the preface (3) — then one begins to see the point! ‘Through Whom unto Thee with Him and with the Holy Ghost’ — but this is the normal introduction of a concluding doxology, a ‘glorifying’ of the Name (cf. Hippolytus /). ‘For great is Thy Name among all nations, and in every place incense is offered unto Thy holy Name . . . For Thou art far above ... every name that is named...’ and so to a climax with the seraphim ‘ever shouting and crying’ as they ‘hallow and glorify’ the dreadful holiness of the Name of God — ‘Holy, holy, holy, Lord of Sabaoth; full is the heaven and earth of Thy glory!’ And then did the people answer, ‘As it was and is and shall be unto generations of generations and world without end. Amen’ — as they still answered at the end of Sarapion’s prayer out of immemorial tradition, though in his day an immense interpolation now divided the sanctus from their response, and his actual ending no longer invited the traditional reply?

We seem to have stumbled on the ‘lost’ doxology of the old Egyptian tradition (cf. p. 172), and a remarkable one it is. But its position carries with it the implication that what follows it, the bulk of the prayer as it now stands — precisely the equivalent in contents of the ‘second half of Hippolytus and Addai and Mari — is an addition to the original nucleus. I do not want to overpress the case, and I will put what appears to me to be the explanation in the form of questions, the answers to which can be weighed by the reader for himself.

In the original Alexandrian prayer was there a series of ‘praisings’ (on the same general scheme as the ‘thanksgivings’ in Hippolytus and other traditions) of which only the first for ‘creation’ now survives, followed by a ‘glorifying of the Name’ with a climax in the sanctus? Is the remainder of the prayer another example of the successive appending of new items in a supplementary position between the original body of the prayer and the communion? (Cf. S. James, pp. 205 sqq., the Lord’s prayer in all rites, the Agnus Dei in the Roman rite, etc.) Is the ‘telescoping’ of the original nucleus (so that the ‘praising for creation’ — its original beginning — now comes immediately before the preface and sanctus — its original ending) a result of the gradual fusion of these supplements with the original eucharistia, and perhaps due to a desire to shorten a prayer becoming unwieldily long by successive additions? Is the strange abruptness which marks the transition from the sanctus to the rest of the prayer in Sarapion c (an abruptness found equally in the transition after the sanctus in S. Mark) explained by the fact that the rest of the prayer was not originally connected at all with the sanctus? (Are the awkward transitions from one section to another throughout the latter part of the prayer of Sarapion to be explained as the marks of successive additions which have never been properly fused together?) Does the phrase, ‘We offer the reasonable sacrifice of this unbloody worship,’ coming where it does in S. Mark (3), explain the original application of the phrase, ‘to Thee we have offered this living sacrifice, this unbloody oblation’ (inserted by Sarapion c at a point after the sanctus) to the angelic worship, as already suggested on p. 166? Have we in S. Mark traces of an original eucharistia of ‘praisings,’ preceded by a ‘Naming’ of God and ending with a glorifying and hallowing of the Name, as the root of the Egyptian liturgical tradition?

Let us now look at the earliest evidence about the contents of the Roman eucharistic prayer, that of Justin, c. a.d. 155. It is worth while studying his language carefully.

1. ‘The president... sends up praise and glory to the Father of all things through the Name of the Son and the Holy Ghost, and makes thanksgiving (eucharistian) at some length that we have been made worthy of these things by Him. And when he has finished the prayers and the thanksgiving (tas euchas kai ten eucharistian), all the laity present shout assent saying ‘Amen...’. And when the president has eucharistised (eucharistesantos) and the people have shouted assent...’ (there follows the communion). (Ap., 1.65.)
2. ‘For we do not take these as common bread or common drink. But as by the Word of God Jesus Christ our Saviour was made Flesh, and had Flesh and Blood for our salvation — so, we have been taught, by a word of prayer which comes from Him, the food which has been "eucharistised" ... is the Flesh and Blood of that Jesus Who was made Flesh. For the apostles in the memoirs which came from them, called "gospels," have recorded that thus it was commanded them — that Jesus took bread and gave thanks and said, "Do this for the anamnesis of Me; this is My Body"; and likewise took the cup and gave thanks and said, "This is My Blood" ‘ (ibid. 66).
3. ‘The president sends up prayers together with thanksgivings (euchas ... eucharistias) to the best of his powers, and the people applaud, saying "Amen"‘ (ibid. 67).
4. (d)... ‘the bread of the eucharist, which Jesus Christ our Lord commanded to be offered for the anamnesis of the passion which He suffered on behalf of men for the cleansing of their souls from all iniquity; that we might at the same time give thanks to God for the creation of the world with all that is therein for man’s sake, and for that He has delivered us from the wickedness wherein we were born, and overthrown the powers and principalities with a perfect overthrow by becoming subject to suffering according to His own counsel’ (Dialogue, 41).

These are the only passages in Justin which appear to deal directly with the contents of the eucharistic prayer (though not the only ones dealing with eucharistic theology), (a) and (c) are obviously summaries of the briefest sort; (b) may or may not refer to something actually found in the prayer as Justin knew it, but the description of the account of the institution as a ‘word’ or ‘ "formula" of prayer which comes from’ Jesus suggests that it had liturgical associations for Justin, (d) is not directly stated to refer to the actual contents of the prayer. But it expresses the meaning of the eucharist, which is what the prayer was intended to do; and it does so in terms so strikingly similar (for a summary) to those of the first part of Hippolytus’ prayer that we need have no hesitation in taking it in this sense.

One might be tempted to infer from Justin’s use of the phrase ‘prayers and thanksgivings’ in (a) and (c) that the eucharistic prayer as he knew it contained an element besides ‘thanksgivings,’ something analogous to the second half of the prayer in Hippolytus. But in view of the order in which he places them, ‘prayers’ before ‘thanksgivings,’ this can hardly be pressed. It might even be argued that in (a) the word euchas ‘prayers’ refers back to the intercessory ‘prayers’ (euchas) before the offertory, mentioned two lines before our quotation begins, where Justin had omitted to mention that the laity replied ‘Amen’ to these ‘prayers,’ an omission which he is now repairing. But the expression a ‘formula of prayer and thanksgiving’ (logoi euches kai eucharistias) is found elsewhere in Justin (e.g. Ap., I. 13) apparently as an elegant variation meaning quite vaguely ‘a thanksgiving to God.’ It seems unwise to assume that he had in mind any rigid distinction in using the two words. In (a) the phrase ‘When he has finished the prayers and the thanksgiving’ is repeated as ‘When the president has eucharistised (given thanks),’ not ‘prayed and eucharistised.’

For the rest one cannot but be struck by the fact that the emphasis in describing the president’s prayer is entirely on the element of ‘thanksgiving.’ It is possible to recognize in the beginning of (a), ‘praise and glory to the Father of all things through the Name of the Son and the Holy Ghost,’ the opening Address and ‘Naming’ of God. At once after this comes ‘he makes thanksgiving ... and when he has finished... the thanksgiving’ the people answer, Amen. So far as the language here goes it would be difficult to say that it suggests any element between the ‘thanksgiving’ and the Amen.

It is quite true that we have already established (p. 159) that there is nothing in the contents of the second half of Hippolytus’ prayer which would not have been accepted by Justin sixty years before him. But this is not necessarily quite the same thing as saying that it was all in the prayer in Justin’s day. It was precisely ideas which were already believed and accepted about the eucharist which people would come to feel ought to be incorporated in the prayer which expressed the meaning of the eucharist. The expansion of the prayer may quite well have taken place in the generation between Justin and Hippolytus, a period about which we know very little, but in which the ideas about the eucharist which they have in common were presumably commonly held in the Roman church. Bating for the moment the question of the institution narrative, which requires separate discussion, all that we can safely say is that Justin’s language is quite consistent with the idea that the Roman prayer in his day consisted only of an Address and ‘Naming’ of God followed by a series of ‘Thanksgivings’ for creation, redemption, etc., and nothing more. If his prayer contained other elements, he has not mentioned them.

As regards the Syrian tradition, we are hampered by a total lack of orthodox documents between Ignatius, c. a.d. 115, and the Didascalia, c. a.d. 250. From Syria we have the Acts of Judas Thomas, which were perhaps composed in the second century. But if so, they have been heavily revised in the third-fourth century, and it is unfortunately the liturgical material which shews some of the Clearest traces of revision. There is, however, a document of the same kind, the Leucian Acts of John, from Asia Minor, which M. R. James was prepared to affirm comes from ‘not later than the middle of the second century.’ We may cite a eucharistic prayer which this puts into the mouth of the apostle, as illustrating at an early stage the eucharistic tradition of Asia which in later times shews more affinities than any other with that of Syria, for which second century evidence is totally lacking.

(a) ‘We glorify Thy Name, which converteth us from error and ruthless deceit:

‘We glorify Thee Who hast shewn before our eyes that which we have seen:

‘We bear witness to Thy loving-kindness which appeareth in divers ways:

‘We praise Thy merciful Name, O Lord.

(b) ‘We give thanks to Thee, Who hast convicted them who are convicted of Thee:

‘We give thanks to Thee, Who hadst need of our nature that should be saved:

‘We give thanks to Thee that Thou hast given us this sure faith,

(c) ‘For Thou art God alone, both now and ever.

It would be very unwise to attempt any reconstruction of the content of the early Eastern eucharistic prayer from this gnostic farrago. But one can detect in most gnostic liturgical practice a steady retention of the orthodox forms while reinterpreting their meaning in gnostic terms and rewriting their formulae in gnostic jargon. Here I draw attention only to the form of this eucharistic prayer. It is addressed not to the Father but to the Son, as is that of Addai and Mari. It opens (a) with a ‘glorifying of the Name’; it consists (b) of a body of four ‘Thanksgivings,’ the number we found in the parallel between the berakah and the second century Roman evidence; and it ends (c) with the statement ‘We give thanks unto Thee, O Holy One’ (hagie), as there is reason to believe that the original Egyptian form ended with a ‘thanksgiving’ (S. Mark 3a) leading up to the ‘hallowing’ of the sanctus. It is fair to say that the same document contains elsewhere (§ 109) another eucharistic prayer in which this structure is less clearly apparent, though it seems at bottom the same.

But it appears safe on the evidence of the prayer above to assert at least that eucharistic prayers of the structure which we have been led to suppose existed in Egypt and at Rome in the early second century were not unknown in the Eastern churches also at that date.

We turn now to what is a more tangled matter, the arrangement of the ‘second half of the prayer as this is found in the various traditions. We are met at the outset by the question, where exactly does this second half begin? There is a broad distinction between the series of Thanksgivings and what follows, but where does the dividing line come? In all the traditions the ‘second half may be defined as lying between an allusion to the last supper (either a full institution narrative or a mere mention) and a concluding doxology. The latter is universal and traceable to the primitive nucleus. Is some reference to the last supper also traceable to this nucleus?

It is difficult to say. On the one hand, such a reference is found in some form in all the traditions. The Jewish berakah in its final thanksgiving for the earthly ‘food wherewith Thou feedest us continually’ contains something which might easily have suggested a thanksgiving for the heavenly food of the eucharist and its method of provision, as the last of the series of Christian ‘Thanksgivings.’ Justin, too, in Ap. I:66, with his formula or 1 "word" of prayer which comes from’ Jesus Himself, suggests that something of the sort stood in the prayer as he knew it.

On the other hand, there are certain difficulties. In all the traditions the reference to the last supper is separated from the Thanksgiving’ series by a sort of intervening clause or ‘link’ (Hippolytus e; Sarapion c; Addai and Mari f). And this link is not the same in any two of them, either in substance or expression. In each case the link itself does not seem at all closely related to the series of ‘Thanksgivings.’ Nor is the allusion to the last supper ever cast in the form of a ‘Thanksgiving,’ but always of a statement. And in the one case where the original ‘glorifying of the Name’ closing the series of ‘Thanksgivings’ has survived in its primitive position (the Egyptian preface and sanctus) the allusion to the last supper comes after this.

This is of some significance. In later times, when the actual history is known to us of the process by which various supplementary items were appended from time to time to the body of the eucharistic prayer between this prayer and the communion, the order in which they are said represents as a rule the sequence in which they were adopted. This is true, e.g., in the Roman rite. The Agnus Dei which was inserted c. a.d. 700 stands before the prayers for unity, etc., which are a still later insertion. We can never quite rule out the possibility of later rearrangement; e.g., in the Roman rite S. Gregory c. a.d. 600 inserted the Lord’s prayer before the pax which had been placed after the canon c. a.d. 400. But the presumption is generally that the earlier additions stand first and the later ones after them. The position of the institution narrative in the Egyptian tradition, both in Sarapion and S. Mark, is that it follows immediately upon the primitive conclusion (the sanctus) with a brief ‘link’ (Sarapion c) between them. This suggests that the institution narrative is originally an addition to the primitive prayer, though an early one, perhaps the very first of all the various items which were appended in course of time to the primitive nucleus of the Egyptian prayer. From the mere position of the institution reference in other traditions one might suspect that the same was true of them also.

But this can hardly be more than a suspicion, even in the case of the Egyptian prayers. One cannot exclude the possibility of a third century rearrangement of the Egyptian prayer when it had already received a certain number of items appended after the sanctus, a rearrangement by which an older reference to the last supper before the sanctus was transferred to a position after it (no doubt with some adaptation) in order to place it in a more central position.

For this much is certain. Whether the reference to the last supper belongs to the primitive nucleus or not, it is the center or pivot of all the developed traditions of the prayer. It serves to cohere the anamnesis of the redemptive work of Christ in the opening series of ‘Thanksgivings’ with the more miscellaneous elements found in the ‘second half’ of the prayer. It is indeed from the reference to the last supper that the substance of this ‘second half grows in every case. In Hippolytus it contains that command to ‘do this for the anamnesis of Me’ which the ‘second half’ goes on to define: ‘Doing therefore the anamnesis ... we offer the bread and the cup,’ etc. In Addai and Mari it is the ‘example,’ which in the Syrian gospel of Matthew contains the promise of that ‘forgiveness of sins’ for which the Syrian churches invariably prayed when they imitated that ‘example’ in their ‘oblation’ (Addai and Mari i). In Sarapion the church does what it does and its offering is what it is because of what our Lord did and said at the last supper: ‘To Thee we have offered this bread, the likeness of the Body ... This bread is the likeness of the holy Body because the Lord Jesus Christ... took bread ... saying . .. "This is My Body" .’

As one reflects upon the great diversity in the ‘second halves’ of these three traditions there appears to be only one likeness of substance between them. Underneath their variety they are at bottom all of them independent attempts to do a single thing, to define the meaning of what the church does at the eucharist and relate it to what was done at the last supper. ‘We offer to Thee the bread and the cup... and we pray Thee that Thou wouldest grant to all who partake to be made one, that they may be fulfilled with Holy Spirit for the confirmation of faith in truth’ (Hippolytus). ‘To Thee we have offered this bread.... We have offered also the cup ... and make all who partake to receive a medicine of life, for the healing of every sickness and for strengthening of all advancement and virtue, not for condemnation ...’ (Sarapion)... ‘this oblation of Thy servants ... that it be to us for the pardon of offences and the remission of sins and for the great hope of resurrection from the dead and for new life in the kingdom of heaven’ (Addai and Mari). This is what the church does at the eucharist — offers and communicates; and it is this which the ‘second half of the prayer expresses and defines. It looks back to the offertory and expresses in words the meaning of that. It looks forward to the communion and prays for the effects of that. The descriptions of the effect of communion are quite differently defined in the three prayers, as can be seen at a glance. The descriptions of the offertory differ verbally more than could have been expected, considering that all three prayers are describing an identical action, of a great simplicity. But essentially they are doing one and the same thing, stating the meaning of the offertory and the communion. It is the function of the prayer to state the meaning of the whole rite.

At this point it may be objected, ‘But what about stating the meaning of the prayer itself and of the fraction?’ Why state the meaning of only the first and last items of the ‘four-action shape’? The fraction was treated primitively as what it had been at the last supper and in the chaburah ritual, a mere preliminary to distribution, without any of the symbolic meanings which were seen in it by later times. And as for the prayer, it was itself the statement of the meaning of the whole rite. A ‘statement of the meaning of the statement of the meaning’ is the sort of refinement which seems to be decisively marked as secondary by mere definition.

Nevertheless the step was taken in course of time, as the churches slowly lost sight of the original principles upon which their rites were framed. And always the statement of the meaning of the prayer is placed between the statements of the meanings of the offertory and the communion. Let us look at two fourth century prayers, from the East and from the West. This time let us take for a change two that we have not hitherto used, those of Apostolic Constitutions, Bk. viii, from Syria, and the Milanese canon cited in de Sacramentis by S. Ambrose, both from the last quarter of the fourth century.

The Eastern prayer runs thus:

a. ‘Making therefore the anamnesis of His passion and death and resurrection and ascension into the heavens, and His second coming that shall be, wherein He shall come to judge the quick and the dead and reward every man according to his works,

Meaning of the offertory

b. We offer unto Thee, our King and God, according to His command this bread and this cup giving thanks unto Thee through Him for that Thou hast made us worthy to stand before Thee and minister as priests to Thee;

c. ‘And we beseech Thee that Thou wouldest favourably regard the gifts that lie before Thee, O God that lackest for nought, and be well pleased with them for the honour of Thy Christ, and send down Thy Holy Spirit upon this sacrifice, the witness of the sufferings of the Lord Jesus, that He (the Holy Ghost) may shew this bread to be the Body of Thy Christ and this cup to be the Blood of Thy Christ:

d. ‘that they who partake of Him may be strengthened unto piety, may receive the forgiveness of sins, may be delivered from the devil and his deceit, may be filled with Holy Spirit, may become worthy of Thy Christ, may receive eternal life, and that Thou mayest be reconciled unto them, O Lord Almighty.’

The Milanese prayer, which is either a ‘first cousin’ or more probably the direct ancestor of the present Roman canon, runs thus:

a. ‘Therefore making the anamnesis of His most glorious passion and resurrection from the dead and ascension into heaven,

b. ‘We offer to Thee this spotless offering, reasonable offering, unbloody offering, this holy bread and cup of eternal life:

c. ‘And we ask and pray that Thou wouldest receive this oblation at Thine altar on high by the hands of Thine angels as Thou didst receive the offerings of Thy righteous servant Abel and the sacrifice of our patriarch Abraham, and that which the high-priest Melchizedek offered unto Thee:’

(At this point the quotation in de Sacramentis ends. But it is virtually certain that the prayer ended much as it ends in the present re-arranged Roman canon):

d. ‘That as many of us as shall receive by this partaking of the altar the most holy Body and Blood of Thy Son may be filled with all heavenly benediction and grace.’

These two prayers each express what is felt as the fundamental meaning of the eucharistic prayer, at the obvious point, between the meanings of the offertory and the communion. The meaning they see in the prayer is different. The Eastern concentrates on ‘consecration,’ the Western on ‘oblation.’ This is typical of a difference which since the fourth century has gradually hardened into a difference of ethos between the Eastern and Western rites and theologies. But it is a mistake to suppose that in the fourth century this distinction had yet acquired a rigidly geographical basis. Mr. W. H. Codrington has recently drawn attention to a whole group of Syrian and Egyptian prayers which contain a reference to the ‘Western’ idea of the offering at the heavenly altar at this point of the prayer. A reference to this same idea is found elsewhere in the rite in Ap. Const., viii. itself, and in the liturgies of S. Basil, S. John Chrysostom and S. Mark. And we must not forget that Sarapion’s prayer is headed ‘Prayer of Oblation,’ even though when it comes to formulate its meaning (in e1) it does so in terms of ‘consecration’ closely allied in thought to those of Ap. Const., viii. c. Similarly it would be easy to find later prayers from Spain and Gaul in the West which state the meaning of the prayer in the ‘Eastern’ way. I am not sure that Ap. Const., viii. c. itself, with its reference to ‘being well pleased with the gifts that lie before Thee,’ is not at least feeling after the ‘Roman’ idea of the oblation at the heavenly altar; while the Roman canon in its turn contains in the Quam oblationem before the institution narrative a petition for consecration expressing the same fundamental idea as the petition in Ap. Const., viii. c, though it is put in quite different theological terms.

Nevertheless, these fourth century statements of the fundamental meaning of the prayer are different. Each concentrates on an aspect of the matter which was clearly recognised from an early date. One has only to remember the phrase of Theodotus in Egypt, c. a.d. 160, already quoted: ‘The bread is hallowed by the power of the Name of God, remaining the same in appearance as it was when it was taken ... it is transformed into spiritual power,’ to see the antiquity of the notion of ‘consecration’ as the chief meaning and purpose of the prayer. On the other hand, one has only to recall the phrase of Irenaeus in the same generation, ‘For there is an altar in heaven, and thither are our prayers and oblations directed,’ to be sure of the equal antiquity of the idea of the heavenly altar at which the eucharist is offered.

But there is another and, it seems, a more penetrating way of regarding this difference of interpretation. In emphasising the meaning of the prayer as ‘consecration,’ is not the one type simply stating in another way the meaning of the communion? And does not the other emphasis on ‘oblation’ only state in another way the meaning of the offertory? In the last analysis the prayer has no separate meaning of its own in the rite to be stated at all. It is not in origin either a ‘consecration prayer’ (in our familiar phrase) or a ‘prayer of oblation’ (as Sarapion called it) but what it was from the beginning — the eucharistic prayer. It is what is ‘done’ at the eucharist, the eucharistic action as a whole, the Shape of the Liturgy, which contains the meaning of the rite. It is the function of the prayer to put this meaning into words.

It is time to draw the threads together. We can distinguish three main periods in the early history of the eucharistic prayer. Working backwards these are:

1. A period in the later fourth and the early fifth centuries, when by a process of mutual borrowing and adaptation all the rites of the great sees are evolving in the direction of a general uniformity of structure and content, and even to some extent of phrasing, in their eucharistic prayers. This is the period which is set up as a norm by the exponents of the traditional theory, who assume that it represents faithfully tendencies which had operated uninterruptedly from the beginning. It is in fact the period which was decisive for the final form of the historic rites. It is represented by such documents as the Roman canon in the West, and S. James, Apostolic Constitutions, viii., and S. Basil in the East (and to some extent by Sarapion, though this is in most respects a document of the preceding period).
2. Behind this is a period covering (? the last quarter of the second century and) the third and earlier part of the fourth centuries. It is marked by the growth of considerable variety in both structure and contents of the unco-ordinated local traditions of the prayer. This is the period upon which the ‘critical’ school of liturgists have fixed their attention. It is represented by such documents as Hippolytus and Addai and Mari (in approximately their present form) and in its later stages by Sarapion and Cyril of Jerusalem. A great deal of work yet remains to be done on the details of the various traditions in this period. But enough is already known for it to be certain that those scholars are right who reject the traditional assumption that the post-Nicene tendency towards uniformity merely developed a pre-Nicene ‘standard type’; or that the Syro-Byzantine outline of the prayer is anything more than one among several amalgams which emerged in the fourth-fifth century. The later fourth century tendency to uniformity was thus a reversal of a third century tendency towards great local diversity. But the critical school in its turn has assumed that the growth of variety in the third century goes back in principle to the very beginning in the apostolic age — so much so that we find Lietzmann and his followers postulating that the eucharistic liturgy never had any single origin at all, but two (or even more) original different sources in the apostolic age.
3. What now of the period behind this again, before the solid evidence of the earliest liturgical texts begins, in the second century and the latter part of the first, which we have been investigating?

The evidence is delicate and scanty, but we seem to have found indications in this period of two distinct strata in the prayer, (a) There are traces of an original stage when the prayer consisted simply of a ‘Naming’ of God, followed by a series of ‘Thanksgivings’ and ending with a ‘hallowing’ or ‘glorifying of the Name.’ This can be connected with the outline of the Jewish ‘Thanksgiving’ which formed an invariable part of that chaburah ritual out of which the ‘four-action shape’ of the eucharist was derived in the latter part of the first century, (b) A second stratum appears to arise out of the reference to the last supper (which may or may not have formed the last member of the original series of Thanksgivings in the first stratum). This second stratum states the meaning of what is done in the celebration of the eucharist, and relates the present eucharistic action of the church to what was done at the last supper.

To me personally the most satisfying thing about the results at which we seem to have arrived is that at no stage of the argument does it require us to go beyond the known facts and the evidence as it stands. We require no silent revolutions accomplished by Antiochene gentile converts, no liturgical innovations by S. Paul, no pagan infiltrations from the mysteries, no inventions or misunderstandings of what happened at the last supper, to account for anything in eucharistic history. And there are no subsequent improbabilities or gaps in the evolution.

That the last supper was a chaburah meeting seems to arise straight out of the New Testament evidence (and indeed from the facts of the case) when this is compared with the ordinary rabbinic regulations for the meetings of such chaburoth. This appears to have been S. Paul’s own understanding of it. The ‘four-action shape’ of the eucharist meets us as an universal fact in the second century. It arises quite naturally from the desire to mark off those particular elements in the chaburah ritual to which our Lord had attached His new meaning, and to separate these from the remainder of the chaburah rite, to which He had attached no special significance. S. Paul’s difficulties at Corinth had foreshadowed the necessity of such a separation, at all events in the gentile churches, long before the end of the apostolic age. The ‘four-action shape’ does in fact detach just these elements from the chaburah rite, leaving the remainder to continue as the agape or Lord’s supper independently of the eucharist.

Among other constituents of the chaburah ritual was the berakah or ‘thanksgiving,’ preceded by a dialogue. Among the constituents of the eucharist was the eucharistia or ‘thanksgiving,’ fulfilling the same function in the Christian as in the Jewish rite, and preceded by the same dialogue-Furthermore, there are traces of a very early stage at which the Christian ‘thanksgiving’ in all traditions had the same outline as the Jewish one, but with the contents rewritten in terms of that ‘New Covenant’ into which it was (according to the earliest tradition) the very purpose of our Lord to initiate His disciples by this rite. So far all is natural, almost inevitable.

Was a direct reference to the last supper included in this primitive eucharistia? It is impossible to decide. One can see very easily why and where it could be placed in the new Christian rewriting of the berakah, and there are things in 1 Cor. 11. (e.g., v. 23: ‘that which I also delivered unto you’) which would make its inclusion from the beginning entirely natural.

On the other hand one must remember the immense difference which the circulation of written gospels must have made to the way in which Christians regarded the historical origin of their faith, and to the store they set by detailed allusions to it. It is extraordinarily difficult for us to think ourselves back behind this change that the written gospels made in the possibility, and therefore the expectation, of such references. But I think I can understand how a gentile Christian late in the first century, introduced to the eucharist for the first time after his baptism, would be content with a tradition that this rite as he found it had been instituted by Jesus, without expecting a detailed account of the institution to be incorporated into the prayer. More particularly would this be the case if his preparation for baptism had not included any biography of Jesus (before the gospels were written or circulating) and not much information about His life beyond the main facts of the crucifixion and resurrection, and some stories of miracles with a number of parables and teachings. (It is, I think, now generally agreed that the primitive preparation for baptism laid emphasis on the Messiahship of Jesus and His atonement, and on moral instructions about conduct, rather than on the history or even the teachings of Jesus in His earthly life.) As for the relation of the eucharist to the chaburah, what gentile convert would understand or care very much about that? It is one of the decisive reasons for placing the formation of the ‘four-action shape’ of the eucharist (which so carefully preserves that relation) right back in the period when even the gentile churches still looked to Jewish leaders, that only jews could have made the changes involved in Jewish custom with such discrimination. And for a Jewish Christian the mere fact that he was now keeping the familiar chaburah ritual with a new meaning, and perhaps with a berakah rewritten in terms of the New Covenant, would be in itself a sufficient reminder of what Jesus was traditionally alleged to have said and done at the last supper, with no need for a specific rehearsing of it. At the most such an allusion as that in Addai and Mari — ‘we have received by tradition the example that is from Thee’ — would suggest itself in such circles.

But once the written gospels came into general circulation (c. a.d. 100-150) even before they were canonised, they would suggest the incorporation into the rite of the sort of account of the institution they contained. The same would be true of the older account in 1 Cor. xi. But one notices that though in later times most rites incorporate other details of S. Paul’s wording, no known rite has the words of institution over the chalice in quite his primitive form, ‘This cup is the New Covenant in My Blood.’ It looks as though all the institution narratives have been suggested by the gospels, even though they fuse them with matter from S. Paul, and treat them in other ways with great independence. I do not see why the incorporation of the institution narrative (or its development from the sort of allusion found in Addai and Mari) should be much later than the period of the first general circulation of the gospels and their public reading in the church, quite early in the second century. This would account for Justin’s description of the words of institution as a formula or ‘word’ of prayer (in Ap. I:66) without difficulty, if it needs accounting for.

The process could hardly stop there, with the mere appending of the narrative to the old Jewish model of the eucharistia. As the church became more and more a purely gentile society and lost contact with its Jewish origins and Jewish habits of thought and ways of piety, the sense of the importance and sufficiency of the Jewish model of the berakah must inevitably fade, and even the understanding of the Jewish basis of the traditional form of the Christian prayer. The idea of the berakah, the series of ‘thanksgivings’ for the work and Person of Jesus the Messiah as in itself an adequate anamnesis of Him before God, had certainly been lost by the churches of the third century, or they would not have overlaid and displaced this Jewish nucleus of the prayer with other elements as they did. Once the historical reference to the last supper had been elaborated or introduced, it provided another focus or center in the prayer. By its mere presence it suggested the need to relate what the church is now doing in the eucharist to this original authority for doing it; and the institution narrative itself contained all the material necessary. ‘Do this for the anamnesis of Me’ — ‘We do the anamnesis of His death and resurrection.’ ‘Take, eat’ — we take and eat, in offertory and communion. There is supplementary matter besides, but that is the framework of the prayer in Hippolytus, our earliest dated text.

The new anamnesis in a sense duplicates matter found in the old Thanksgivings, but with a different emphasis. The old matter concentrates on the Person of Christ — it is an anamnesis of ‘Him’ — and on the effects of redemption. The new anamnesis derived from the historical narrative of the institution concentrates on the particular events in history by which redemption was wrought — ‘His death and resurrection.’ We have already noted that Hippolytus e (the introduction to the institution narrative) regards the eucharist as the present means by which these ‘effects’ of redemption are actually achieved in the individual soul. Thus the institution narrative has drawn to itself before it the essence of the old ‘Thanksgivings,’ just as it furnishes the basis for the whole second half of the prayer. It has become the focus or pivot of the whole, linking the old and the new material. And very rightly, for it contains in itself all that our Lord had said as to the new meaning to be attached by his followers to ‘doing this,’ the very pith of that meaning of the rite which it was the function of the prayer to state.

The development in other churches was not quite the same; there is, e.g., no anamnesis in Sarapion, nor, I think, originally in Addai and Mari. But everywhere there is the manifest intention that the second half of the prayer should state the meaning of offertory and communion in relation to the last supper. Everywhere the second half of the prayer has its roots in the allusion to the last supper, even though it was the eucharistic action, the Shape of the Liturgy from offertory to communion, that provided the substance of this part of the prayer.

The question arises as to the date when this development of the institution narrative into the ‘second half’ of the prayer may be called an accomplished fact. Where all opinions are bound to be tentative I can only put the matter as it seems to me. Hippolytus c. a.d. 215 is a terminus ad quem. More than one scholar has recently questioned whether the prayer as it now stands in the text of the Apostolic Tradition has not been interpolated since his day. On grounds of textual criticism I believe this suspicion to be true of one clause in Hippolytus k. But for the rest the textual tradition is astonishingly unanimous as to the substance in versions in Latin, Greek, Syriac and Ethiopia And there is this further consideration: Hippolytus is a writer with a strongly marked personal style and vocabulary, who is much given to repeating little tags or catch-phrases of his own. Almost every clause in the prayer as it stands can be paralleled in style, vocabulary and even phrasing, some of them many times over, in other unquestioned works of his. And these parallels, some of which have been collected by Dom Connolly, are found in all parts of the present text. The prayer as it stands may be taken as coming from his pen — more than that, as being of his composition. I mean by this, not that he is the inventor of this type of prayer, but that its phrasing and articulation bear unmistakable marks of his personal ideas.

In the circumstances in which the Apostolic Tradition was issued — as a conservative manifesto against contemporary innovations in the Roman church — we must attach a good deal of weight to Hippolytus’ claim that he is setting down customs which had been traditional at Rome at least during his whole life-time, say from c. a.d. 175 or rather earlier. But this must not blind us to the fact that there are a number of phrases in the prayer which are distinctive of his own peculiar theology of the Trinity, and which the rest of the Roman church in his own lifetime might very well have refused to use. Yet the general form and structure of the prayer are very unlikely to have been unusual at Rome in his day. It would have stultified the whole purpose of his pamphlet in favour of the old ways if the first prayer he gave as an example was of a type unknown to the average Roman Christian, or even one which his Christian contemporaries would not recognise as like those in use there ever since they could remember. But that very ‘tidyness’ and closeness of articulation which distinguish his prayer from those of Addai and Mari and Sarapion are a sign that in the prayer of Hippolytus the material has been thoroughly fused and ordered by a single mind. It is the product, on strictly traditional lines, of a professional theologian. In Addai and Mari and Sarapion we have the much less orderly and coherent result of the gradual accumulations of tradition in local churches.

Nevertheless, Hippolytus supplies us with a lower limit which we can accept with some confidence. The eucharistic prayer at Rome had had some sort of ‘second half’ ever since he could remember — say since c.a.d. 175. If the evidence of Justin is to be taken at its face-value, the Roman prayer had been expanded to include this only in the preceding quarter of a century. Development in some churches may have been less rapid, but it may well have been more so. Even in the second century the Roman church had deserved a reputation for conservatism.

The theory sketched here of the second century development of the prayer will probably seem to many impossibly radical. I can only plead against the traditionalists that the actual structure of the prayer in all traditions suggests that in its simplest form it contains two separate strata,’ that the ‘Thanksgiving series’ and the ‘second half’ spring from two different roots, serve two different purposes and are fused into a single prayer only by the allusion to the last supper. Even if I am wrong in supposing that the ‘second half’ is a later addition — and I have tried to shew that there is definite historical evidence to be discerned for thinking that it is — the construction of the prayer itself would still oblige us to believe that it was originally framed as two halves and not as a unity.

It is equally likely that liturgical experts who accept the theories of Lietzmann and his school will see here chiefly a return to the essential point of the traditional theory — the single origin of the eucharistic rite. This involves the rejection of that original ‘duality’ which scholars like Ceriani and E. Bishop avowed that they found in early liturgical history, and which their modern successors have traced to a fundamental division in eucharistic doctrine and practice between S. Paul and the Judaic apostles headed by S. Peter. I must answer plainly that in its modern contemporary form this theory is only one more of those visitations by the ghost of F. C. Baur to which theological scholarship is still occasionally liable. The Tubingen romance of an apostolic schism is no more soundly based in the early history of the liturgy than it is in any other branch of church history. Its survival among liturgists after it had been discarded as untenable by historians (with the exception of Lietzmann himself in his Beginnings of Christianity, E.T., 1937) has been the principal hindrance to the progress of liturgical studies in the past twenty years.

S. Paul was a jew and a rabbinic student and a pharisee. Like the Jewish church before him he used a thoroughly Jewish rite at the eucharist, as did the Pauline churches after him. That was inevitable. The ‘Pauline’ eucharist arose at Jerusalem, from a new meaning given to something authentically, integrally, traditionally Jewish, the chaburah meal of the last supper. I have set out the evidence, and by that every theory in the end must stand or fall. But I claim that on the evidence it is right to assert that in the last analysis Frere and his predecessors were right, as against Lietzmann and his followers and predecessors, in attributing to the eucharist and the liturgy which performed it a single origin, and not a dual one. Their failure lay in a refusal to pursue the question to its roots, and to insist that if there was such a single aboriginal type of eucharistic prayer, it must in the nature of the case have been on a Jewish model and not on a Greek one. Developed in this direction the traditional theory of a single origin to the liturgy ‘fits’ the evidence at every point as the theory of a dual origin has never fitted it because it is not true.

Certainly there was a duality — it might be truer to say a plurality — about the interpretation of the eucharist from the beginning. One can trace it even in the New Testament.1 But it is a multiplicity of meanings seen in a single action. That action was one and fixed from the evening of the last supper — ‘do this’ — and the rite that ensured its perpetuation was one and fixed in its form so far back as we can trace. What grew — as our Lord meant it to grow — and broadened and deepened and enriched itself in ever new ways as the Christian generations passed was the meaning drawn from the words ‘for the anamnesis of Me’ (cf. p.4).

One result of the fourth century transformation of the eucharist into a fully public act is a certain elaboration of ceremony in its performance. This does not directly concern the subject of this book, since the Shape of the Liturgy by which the eucharistic action is performed is hardly affected by this. The introit-chant, which covered the processional entrance of the clergy, seems to be the only item in the outline of the rite which was introduced for purely ceremonial reasons. The eucharistic action and its meaning remained in themselves what they had always been. But the actual performance of the action is to a certain extent formalized in a new way in the fourth century. There is a new emphasis on its earthly and human aspect, consistent indeed with the acceptance of a mission to human society as such and that sanctification of social living in time which the church first undertook in the fourth century, but also a symptom of the decline of the old eschatological understanding of the rite.

Yet here also the Constantinian and post-Nicene church made no deliberate breach with the past and was quite unconscious of any new beginning. As we have seen, from the very fact that it was a corporate action the pre-Nicene eucharist had had an aspect of ceremony ever since the first formation of the liturgical eucharist apart from the supper, in that it required a good deal of concerted movement by all the various ‘orders’ of participants for its performance. This core of the action, which was everywhere the same," is in its origin wholly utilitarian — it is the simplest and most natural way of getting the corporate eucharistic action ‘done.’ But by the fourth century it had already hardened into something very like a traditional ritual by the mere passage of centuries. The post-Nicene church had obviously every intention of conserving this pre-Nicene body of customs intact, and it does in fact form the whole basis of the later eucharistic rites. But it soon began to be overlaid and accompanied by a variety of new customs. Some of these, like the solemn processional entry of the clergy at the beginning of the rite, were suggested quite naturally by the new public conditions of worship and its more formal setting. Others, like the lavabo, were deliberately symbolical, and intended to remind the worshippers in various ways of the solemnity of what they were about. These may be innovations, but they seem natural products of the new situation. Now that not only the spiritually sensitive but the average man and woman were increasingly becoming regular attendants at Christian worship, the introduction of such reminders of its solemnity was a necessary part of the church’s care for her members. The Reformers of the sixteenth century, who regarded the eucharist primarily as something ‘said’ by the clergy, set themselves to achieve exactly the same object by prescribing solemn and lengthy ‘exhortations’ to be said by the minister to the worshippers (of which specimens still remain to us in the ‘Long’ and ‘Short Exhortations’ of the Prayer Book rite). The fourth century church took more literally the command ‘Do this in remembrance of Me,’ and therefore addressed such reminders to the people by symbolical gestures and actions rather than by words. But the purpose in both cases is exactly the same. It is the change made by regarding the rite as something ‘said’ and not something ‘done’ (which is essentially the work of the Latin middle ages and not of the Reformers) that makes it difficult for modern Western Christians, Protestant and catholic alike, to enter immediately into the mind of the early church.

Mr. A. D. Nock in his brilliant study of the psychological process behind the conversion of the pagan world to Christianity has remarked that ‘Even in the fourth century, when the Eucharist acquired a dignity of ceremonial appropriate to the solemn worship of the now dominant church, it is not to me clear either that there was a deliberate copying of the ceremonial of the mystery dramas or that any special appeal was made by the ritual to the mass of new converts.’ I venture to hope that what has been already written is sufficient comment on the question of possible copying of the mysteries in pre-Nicene times. We have seen that there is in fact no element in the eucharistic ceremonial, such as it was, of the first three centuries which is not completely explained by a directly Christian or pre-christian Jewish origin. As for any appeal of ceremonial to the fourth century converts, there is nothing in the evidence which suggests that this was its intention. The eucharist was now being performed in a world where every public act secular or religious had always been invested with a certain amount of ceremony as a matter of course. Christian worship was now a public act, and any different treatment of it was simply not thought of. A few notes on the chief adjuncts of ceremonial and their introduction and development will make clear, I think, how spontaneous the whole process of the post-Nicene development of ceremonial really was.

What one may call ‘official costume for public acts’ both in the case of magistrates and priests had been common in classical Greece and usual all over the Near East for many centuries before the Christian era. In Italy and the West, particularly at Rome, the wearing of such ‘official’ robes, either secular or religious, had always been much less developed; though the elements of the idea are to be discerned, e.g. in the toga praetexta of the magistrates (the ordinary dress of a gentleman, with a broad purple stripe) and the apex — the special skin cap worn on some occasions by the Roman pontiffs and flamens. But speaking broadly, the elaborately vested Maccabean high-priest performing the rites of the Day of Atonement on the one hand, and on the other the pagan Roman rex sacrorum performing the not very dissimilar rites of the poplifugiwn in the toga which every Roman gentleman wore about the city every day, represent from one point of view a contrast of types whose basis is geographical much more than dependent on different ideas of worship.

It is therefore not surprising to find that the earliest mention of a special liturgical garment for use at Christian worship comes from the Near East, and specifically from Jerusalem. We learn incidentally from Theodoret that c. a.d. 330 Constantine had presented to his new cathedral church at Jerusalem as part of its furnishing a ‘sacred robe’ (hieran stolen) of gold tissue to be worn by the bishop when presiding at the solemn baptisms of the paschal vigil. From the words employed this looks like some sort of special liturgical vestment. But this very characteristic initiative of the ritualistic Jerusalem church was not followed up. The next mention of such things comes likewise from Syria, in a rubric of the rite in Ap. Const., viii. (c. a.d. 375) directing that the bishop is to celebrate the eucharist ‘clad in splendid raiment.’ But the word estheta in this case makes it clear that all the author has in mind is a sumptuous specimen of the ordinary lay costume of the upper classes at this period, not a special hieratic vestment (stole) like those of the Old Testament high-priests. And in fact the Roman type of sacerdotal functioning in ordinary dress did prevail in Christian usage everywhere over the graeco-oriental type of a special liturgical dress. All over Christendom ecclesiastical vestments derive from the lay dress of the upper classes in the imperial period, and not from any return to Old Testament precedents such as the mediaeval ritualists imagined.

The Chasuble, Tunicle and Alb. Since the second century the old Roman toga virilis had been more and more disused as an everyday garment, and was no longer worn even at ordinary meetings of the senate. In place of the toga the upper classes adopted a costume, apparently Ionian in origin, consisting of a linen robe with close sleeves, covering the whole body from neck to feet, the linea, above which was worn a sort of tunic with short close sleeves (colobium or tunica) extending to the knees. On formal occasions and out of doors both men and women wore over this the paenula (also called planeta, casula and occasionally lacerna) — a large round piece of stuff with a hole in the centre for the head to pass through, which fell in folds over the shoulders and arms and draped the whole body down to the knees.

The contemporary account of the martyrdom of S. Cyprian in a.d. 258 reveals him as wearing this dress. When he reached the place of execution ‘he took off the red lacerna that he was wearing and folded it and knelt down upon it and prostrated himself in prayer to the Lord. And when he had taken off his tunica and handed it to the deacons, he stood up in his linea and awaited the executioners.’ These are in essentials the pontificals of a mediaeval bishop. But Cyprian is wearing them simply as the ordinary lay gentleman’s dress of the day.

By the end of the fourth century this peaceful costume in turn was beginning to go out of fashion in favour of a more military style brought in by the barbarian mercenaries whose commanders were becoming the most influential people in the state. By a law of a.d. 397, however, senators were ordered to resume the old civilian style of the paenula worn over the colobium or tunica and the ungirded linea,’ while civil servants were ordered to wear the paenula over the girdled linea as part of their full dress. (The cingulum (belt) was a distinguishing badge of military as opposed to civil office. Hence the officiates, whose service ranked as a militia and was subject to military not civil law, are to wear the girdle visibly, but the senator, as a civilian, does not). In the rigidly organised late empire this law sufficed to fix the costume of the great nobles and the higher officials. Two centuries later, in the apparently contemporary portrait of Pope S. Gregory I standing between his father the senator Gordianus and his mother, the costume of all three is still exactly the same — chasuble worn over the tunic with the ungirded linen alb. The mother wears a sort of linen turban, and the Pope is distinguished from the layman his father by the pallium — a sort of scarf of office which was the only strictly liturgical vestment which the Popes as yet tolerated. But otherwise the costumes of the bishop, the layman and laywoman are exactly the same.

The Pallium and Stole. Even the use of the pallium was not very ancient in the Roman church, dating perhaps from the end of the fifth century. Before that time the whole idea of any such mark of distinction had been entirely contrary to the local Roman tradition. Pope Celestine I c. a.d. 425 had gone so far as to rebuke the bishops of the South of France, among whom the use of the pallium and girdle at the eucharist was already customary, with what seems unnecessary vigour: ‘It is small wonder that the church’s custom should be violated by those who have not grown old in the church, but entering in by some other way have introduced into the church along with themselves things which they used to wear in another walk of life (i.e., the magistracy, from which so many bishops were then recruited).... Perhaps men who dwell in distant parts far from the rest of the world wear that dress from following local custom rather than reason. Whence came this custom in the churches of Gaul, so contrary to antiquity? We bishops must be distinguished from the people and others by our learning not by our dress, by our life not by our robes, by purity of heart not by elegance .., To the plea that this is only a literal following of the evangelical injunction to have ‘the loins girded,’ etc., he answers drily that they will need to stand at the altar with a burning lamp in one hand and a staff in the other to fulfil what follows, and roundly bids them to have done with such ‘worthless superstitions.’

Yet there is evidence from the East as well as from Gaul that in other churches less sturdily old-fashioned than that of Rome some equivalent of the pallium had already been accepted as a special badge of the liturgical ministry almost everywhere during the later fourth century. It is in fact the liturgical ‘vestment’ (stole) of all orders at this time. In its episcopal form the pallium is simply the old ‘scarf of office’ worn by the emperor and consuls, a badge granted to numerous other officials during the fourth century. It was adopted by the clergy in various forms, becoming the pallium of the Pope and (later) of archbishops and certain privileged bishops in the West, but worn by all bishops since the fifth century in the East. For the lower clergy it becomes the ‘stole’ worn in different ways by bishops, priests and deacons as a badge of distinction. Most pallia, lay and clerical alike, were of coloured silk. But the Popes when they adopted this little piece of vanity wore it in the form of a simple white woollen scarf embroidered with black crosses. And apart from the Pope’s pallium Rome so far remained faithful to Celestine’s principles as not to adopt the stole in any form, for bishops, priests or deacons, right down to the twelfth century, when it was introduced from beyond the Alps.

The Maniple. Just as the pallium and stole derive from the secular ‘scarf of office,’ so the vestment known as the maniple (fanon, sudarium) derives directly from the mappula, a sort of large handkerchief which formed part of the ceremonial dress of consuls and other magistrates, carried in the hand or laid across the arm. The carrying of the maniple in the left hand at the liturgy did not die out at Rome or in England before the twelfth century, though by then the present custom of fixing it to the left arm throughout the rite was firmly established in France. The use of the mappula by the clergy is attested at Rome in the sixth century, but it is found as a special badge of the deacon in Egypt a century before this.

The Dalmatic. This was a form of tunica with large sleeves, which came into use in the second century as a tunic which could be worn in public without the chasuble (though it was noted as a breach of decorum in the emperor Commodus that he appeared sometimes at the circus clad only in the dalmatic without the chasuble). In the fourth century it seems to have become a sort of undress uniform for high officials, and as such it began to be worn by important bishops, though always under the chasuble. It was adopted by itself, however, as a normal dress by the seven regionary deacons of Rome, whose duties, as superintendents of what was now virtually the whole poor relief system of the city (pauperised for centuries by the system of panis et circenses) and the estates which formed its endowment, were becoming administrative and financial rather than religious. For a while this remained a peculiarity of the Roman deacons, but it spread gradually to other Western churches, where it eventually, became the distinctive vestment of deacons. It is symbolic of a good deal in church history that the adoption of this dress, which was virtually a badge of preoccupation with secular affairs, was at Rome confined to the deacons, while in the Byzantine church it became the special vestment of archbishops. Even the Roman deacons, arrogant and worldly as a long series of critics from the fourth to the sixth century declared them to be, hesitated to perform their liturgical functions in this uniform of a secular official. The Pope, who as the chief citizen of Rome sometimes wore a dalmatic, always covered it in church with the chasuble of the private gentleman. The deacons at least began their ministry at the altar dressed in the same way. But before performing his special ‘liturgy’ of singing the gospel (and down to a.d. 595 the preceding solo of the gradual) the Roman deacon put off his chasuble, which he only resumed after assisting to administer communion. There was no mystical or symbolic meaning in this; it was simply for convenience of movement. In Lent the Roman deacons acknowledged the special seriousness of the season by leaving off their dalmatics in church and wearing their chasubles throughout the rite. But even so they wore them from the gospel to the communion rolled up bandolier-wise around the body over the left shoulder and tied under the right arm — something like a British soldier’s greatcoat in the period of the Boer War. (This curiously informal behaviour is still perpetuated in the ceremonial of the Roman rite in Advent, Lent and Ember-tides.)

The Camelaucum or Tiara, It is the same story with the other vestments that originated before the middle ages. The Papal tiara, for instance, is derived from the camelaucum or phrygia, a ‘cap of state’ worn by the emperors and very high officials in the fourth century. (The statue of Constantine on his triumphal arch at Rome is wearing one. A version of the same headgear was worn by the doge of Venice and other Italo-Greek potentates.) Its use seems to have been allowed to the clergy by the emperors everywhere in the fifth century. In the East, in the form of the ‘brimless top-hat’ doubtless familiar to most readers, it became the normal headgear of all clergy (white for patriarchs like that of the Pope, purple for bishops and black for others). Like the Western biretta, it began to be worn by Easterns in church as well as out of it during the later middle ages. Down to the tenth century the Popes kept it as a strictly non-liturgical vestment, to be worn to and from church and on other public occasions, but not in service time like the later mitre, though the latter seems to have evolved from it by a process of variation. When the Popes became secular rulers in their own right (from the ninth century onwards) they successively added the three crowns (the last was added in the fourteenth century) to the camelaucum as a secular headgear, but they have never worn this crowned camelaucum, the ornament of a secular ruler, while celebrating the eucharist.

The Campagi or Shoes. The special liturgical shoes and stockings of Western bishops also originated as a secular ornament, worn outside church as well as at the liturgy. As far back as the early days of the Roman republic consuls and triumphing generals were distinguished by high-laced shoes of a particular form and a bright red colour; and patricians were distinguished from plebeians by a particular form of black shoe. In the fourth century a.d. when all dress was formalised and regulated with a sort of childish care into nicely distinguished badges of rank, the wearing of different forms of shoes by different orders of officials was a matter for imperial edicts. The purple boots of the Byzantine emperors became, like the purple chasuble embroidered with golden bees, the most jealously guarded symbol of imperial power, even more so than the diadem. To assume them was to claim the throne. In adopting the campagi as part of their liturgical dress, probably in the fifth century, the Popes were only carrying out their customary policy of celebrating the liturgy in the normal dress of important laymen of the time. But by the sixth century the campagi like the pallium must have come to be reckoned a distinctive sign of their episcopal office.

These are the only ecclesiastical vestments worn in Christendom before c. A.D. 800. In their adoption there is evidence of a definite policy pursued everywhere during the fourth and fifth centuries, viz., that the liturgy should be celebrated always in the garments of everyday life. The use of symbolical liturgical vestments like those of the Old Testament priests or the white dress of the neophytes after baptism in the pre-Nicene church was deliberately avoided. The only exception, if it can be called such, was the introduction of the stole; but scarves of office of all kinds were so commonly used in social and civic life in the later fourth century that this, too, can be brought under the same heading, even though Rome thought otherwise and refused to adopt it for seven centuries or so, except for the bishop.

What turned this clothing into a special liturgical vesture was mere conservatism. When the dress of the layman finally changed in the sixth and seventh centuries to the new barbarian fashions, the clergy as the last representatives of the old civilised tradition retained the old civilised costume. From being old-fashioned it became archaic (like the court-dress of the Moderator of the Scottish Kirk) and finally hieratic (like the chimere of the Anglican bishop, which begins prosaically in the twelfth century as a form of overcoat).

But this last stage was only reached by degrees, and was not complete before the seventh-eighth century. Where the old tradition lingered amongst the laity, there the old dress lingered for laymen too, as we see from the picture of S. Gregory and his father Gordian, c. a.d. 600. But though the old-fashioned patrician families of Rome might preserve the traditional dress in everyday life, elsewhere it had already vanished. The fourth Council of Toledo in a.d. 633 orders the public restoration before the altar of the chasuble, stole and alb to an unfrocked priest who is being restored to the use of his orders — a provision which tells its own story; the old costume has become a strictly clerical vestment, a liturgical symbol. The Byzantine emperors continued to wear it in proud assertion of their claim to continue in unbroken succession the office of the Roman Caesars. In 1453 the last emperor of Constantinople fell in the breach fighting to the end, still clothed in the purple chasuble embroidered with golden bees. Charlemagne adopted this with the purple buskins when he laid claim to the imperial dignity in a.d. 800; and from him it passed to the kings of France as their coronation robe. The last public use of it by a layman was at the coronation of Charles X of France at Rheims in 1825.

We can better understand the process if we compare it with the history of a garment with which we are more familiar. A century ago the black frock-coat was still the dress of every Englishman above the condition of a labourer. Even forty years ago a large proportion of the upper classes wore it on Sundays and on any occasion of formality. Now it is gradually becoming an undress uniform for royalty, diplomats and statesmen, and for people in certain formal positions, shopwalkers, undertakers, important station-masters — and Anglican dignitaries. Even bridegrooms had abandoned it for the morning coat before the war. It was adopted for use in conducting divine worship by many non-conformist divines in the last century, precisely because it was the normal lay dress of the time. But many of them retain it to-day when it has ceased to be so, and their people would be mildly shocked by a change. One delightful old Baptist lay-preacher whom I knew in Pembrokeshire nearly thirty years ago always referred to it as his ‘preaching coat,’ and would never have used it for any other purpose. It is on its way — just like the chasuble — to becoming a vestment, a special royal and liturgical garment.

The case is, however, quite different with the vestments which developed later, the Mitre, Cope and Gloves, and the choir dress of Surplice, etc. These mediaeval vestments were of deliberate clerical invention, and were meant in their ecclesiastical form to be worn only at the liturgy, and as clerical marks of distinction from the remainder of the worshippers.

The Mitre. The advocates of an inner connection of the catholic eu-charist with the pagan mysteries have had interesting things to say in the past about the episcopal mitre, the headgear whose very name recalls the hierophant of Mithras. It is unfortunate for such theories that the mitre (mitra, mitella) first appears in Christian use as the distinctive headgear of the only person who had no particular function in the liturgy — the deaconess. References to its use by deaconesses in Africa are found in the later fourth century. It passed thence to Spain where a seventh-eighth century mention of the mitra religiosa in the form for the installation of an abbess (reckoned ex officio a deaconess) is preserved in the Mozarabic Liber Ordinum.

The German scholar Pater Braun in his exhaustive discussion of every piece of evidence which has ever been alleged for the antiquity of the episcopal mitre appears to have proved decisively that no liturgical headgear whatever was ever worn by the clergy at the liturgy anywhere before c. a.d. 1000. The change in this comes during the eleventh century in the West. The first mention of an episcopal mitre in literature is the grant on Passion Sunday a.d. 1049 by Pope S. Leo IX to his own former archbishop Eberhard of Trier of the right to wear ‘at the liturgy’ (in ecclesiastids officiis) ‘the Roman mitre,’ ‘after the Roman fashion.’ In 1051 the Pope grants the same privilege to the seven ‘cardinals’ (i.e. principal chaplains) of the cathedral of Besancon when acting as celebrant, deacon or sub-deacon at the high altar on certain great feasts. This privilege of wearing mitres at the liturgy was granted to a number of other chapters of canons (even for their subdeacons) during the next half century or so, sometimes on the occasion of the grant of a mitre to their bishop, sometimes actually before this. In 1063 the mitre was granted to Abbot Elsin of S. Augustine’s, Canterbury (the first of many such grants to abbots); and though Braun takes it for granted that this proves that the mitre had already been granted to his archbishop, Stigand’s pontificals in the Bayeux Tapestry (which are very carefully portrayed) do not include the mitre at Harold’s coronation.

Great churches like Milan only obtained the privilege of the mitre at die beginning of the twelfth century, and it was not until the middle or third quarter of that century that it came about that so many bishops had acquired the right to use it by specific Papal grant that it began to be regarded as an inevitable part of a bishop’s costume, and the remaining non-mitred bishops simply usurped it without obtaining a Papal grant. Abbots, conventual priors and other dignitaries continued to obtain it individually by a privilege from the Pope in the old way until the seventeenth century, when the few remaining non-mitred abbots were granted the use of it ex officio.

The real origin of the liturgical mitre would therefore seem to be as follows: We know that in the tenth century the Popes still did not wear their camclaucum at the liturgy. But somewhere soon after a.d. iooo they must have begun to do so, differentiating however between this use of it and that outside church by reserving the ‘crowned’ camelaucum (for the first of the three crowns had by now been added to the papal cap) for secular occasions. It is this new use of the camelaucum in church which is allowed to Eberhard of Trier; and the grant to the cardinals of Besancon in 1051 suggests that it was already used in church by the Roman cardinals before this. The mitre is thus the one and only liturgical ornament of purely Papal origin; and the right of others to use it, whether bishops or priests or deacons or subdeacons or even laymen (some mediaeval princes, e.g. the kings of Hungary and some dukes of Bohemia, were granted this as a compliment), depended originally on a Papal privilege even more strictly than did the use of the pallium. It is in no sense a symbol in itself of episcopal orders, even though it is now worn by all Western bishops, including Swedish Lutherans and others who are not in communion with the Pope. Apart from Papal initiative it would have remained an ornament not of the bishop at all, but of the deaconess.

The Eastern mitre, in the form of a crown, has a wholly different origin. It seems to derive from the touphan, a sort of jewelled turban borrowed by the Byzantines from the Persians. But its use by ecclesiastics in church is not older than the sixteenth century. The great Byzantine canonist Balsamon states categorically c. a.d. 1200 that all Eastern ecclesiastics are bare-headed at the liturgy with the sole exception of the Patriarch of Alexandria and his twelve ‘cardinary’ priests, who wear a loron (diadem), a right which he says was acquired by S. Cyril as the Papal legate at the Council of Ephesus in a.d. 432. The same statement is twice repeated by Simeon of Thessalonica in the fifteenth century.

The Cope, of silk or velvet and embroidery, is an elaboration for the deliberate purpose of ecclesiastical display in church of the homely cape for keeping warm. It was invented in the great French capitular and conventual churches during the ninth century, and was in occasional use for semi-liturgical functions (e.g. the dialogue at the Easter Sepulchre) in England in the later tenth century. It was still not in use at Rome in the twelfth century.

The Episcopal Gloves. The use of gloves does not seem to have been known in antiquity at all. They first appear as episcopal ornaments in Gaul during the ninth century and were adopted at Rome during the tenth or eleventh century. A trace of their late origin in episcopal costume is to be found in the fact that in the Western Pontificale they are assumed by a new bishop not when he puts on the old pontifical vestments at the moment of his consecration, but only after the communion (like that other afterthought the mitre, which is placed upon his head at the same point of the rite, after all that relates to his episcopal consecration has been concluded). Their liturgical name chirothecae suggests that they came into the Latin churches from the Greek countries; but the liturgical use of gloves (properly so called) is unknown in the East. Byzantine court-dress, however, included a pair of embroidered cuffs (epimanikia, manualia), which appear among the vestments of most Eastern rites. These were borrowed from Byzantium by some Spanish and French churches in the eighth century. They may be regarded as embryonic gloves. At all events, their use in the West as episcopal ornaments went out when that of gloves came in.

Choir-Dress. The Surplice. At the eucharist all the clergy down to and including the acolytes (with the partial exception of the deacons) wore the chasuble in the fifth century; and traces of this practice continued at Rome down to at least the ninth century. But clerical dress at the divine office, at all events in the case of the lower clergy, seems to have been always the girdled linea, or alb, the ‘undress’ of the middle classes at home. It was not a very warm costume, and the difficulty of heating the church, especially for the long night office, was solved by heating the man instead. Thick fur coats (pelliceum) worn under or over the alb were a necessity. The awkwardness of such bundlesome garments under the girdled alb led to the disuse of the girdle, and the surplice (superpelliceum) is simply the alb adapted for use ‘over the fur coat.’ The graceful flowing sleeves of the mediaeval surplice seem to have been added early in the thirteenth century, as part of the deliberate beautifying of all church vestments which is a noticeable feature of that period. Before that time the comparatively close sleeves of the cotta (then still a garment which came below the knees) preserved more nearly the original resemblance to the ungirded alb.

The Rochet is simply the alb or linea retained as a secular dress by the clergy for use outside church. It is an unliturgical garment, over which both priests and bishops were perpetually being reminded by mediaeval synods that they ought to assume the surplice whenever they had to perform any properly ecclesiastical duty whatever. It retains its character as a secular dress for the clergy in the Church of England, as the distinctive robe of bishops at sittings of the House of Lords. As a semi-clerical but obviously non-priestly dress it was recommended by many councils as a suitable garb for sacristans and sextons, and for laymen who had not received even the first clerical tonsure when serving mass. Its origin as a properly liturgical vestment appears to lie in its tolerated use after the middle of the thirteenth century by parish priests for administering baptism, when the new long sleeves of the surplice were liable to trail in the font.

Apart from this, its use in church as a distinctive garment for prelates and dignitaries has a slightly unedifying origin. It appears that in the late twelfth century the canons of S. Peter’s at Rome got into the way of not troubling to put on the surplice over the rochet (which they still wore as part of their out-door dress) for the daily recitation of the office in church. Ignorant copying of this slackness by foreign prelates visiting Rome set a new fashion, which by the fifteenth century had hardened into a general custom; though the rubrics of the liturgical books have never yet ceased to require the use of the surplice over the rochet by dignitaries for even the most trivial liturgical duties. The fastening of this little piece of mediaeval Italian slovenliness upon all Anglican bishops by the rubrics of the Prayer Book of 1552 is one of those curiosities of liturgical history which add at once to its interest and to its complications.

The various forms of almuce, mozzetta, hood, tippet, scarf, etc. are all mediaeval or later. They are all derived ultimately from the fur coat or cloth cape worn over the linea (instead of under it) for warmth. They are formalized in various ways (shape, colour, material) partly as badges of rank and distinction amongst the clergy themselves, partly in order to distinguish the ordained from the unordained cleric when all alike are wearing the surplice.

The Eastern church has never developed a choir-dress for the secular clergy, chiefly because since the seventh or eighth century the Eastern secular clergy has abandoned the regular recitation of the office to the monks, who like all monks recite it in their habits. When the oriental clergy do conduct parts of the office in public they wear their eucharistic vestments, as at the administration of all sacraments. This would seem to have been the practice of Western clergy, too, before the invention of the cope and surplice.

This review of the history of vestments, though sketchy, is sufficient to establish two main points:

1. That in the fourth century, as before, the ‘domestic’ character of early Christian worship asserted itself even after the transference of the eucharist to the basilicas sufficiently to prevent the adoption anywhere of special ceremonial robes, such as were a usual part of the apparatus of the pagan mysteries. There was indeed no intention whatever of setting up any distinction of dress between clergy and laity at the liturgy. (The adoption of the stole would find its modern equivalent, I suppose, in something like a clerical collar, or a steward’s rosette at a secular meeting.) 2. That by the beginning of the middle ages such a distinction had grown up accidentally by the mere fact that the clergy in church retained the old universal costume after the laity had discarded it. The idea of a special liturgical dress for the clergy came then to be accepted as something right and desirable in itself — an idea which has persisted. For it is to be noted that the adoption by the minister of a Geneva gown and preaching bands, or of a surplice and academic hood, is as much the adoption of a special liturgical costume as the use of eucharistic vestments. For that matter in these days the Salvation Army’s poke bonnet and the black frock coat with a white bow tie follow the precedent of the mitre and pallium, not that of the chasuble and dalmatic, in that the use of these things is deliberately intended to distinguish the wearer from his or her fellow Christians at the liturgy; whereas the older vestments were originally intended to do exactly the opposite.

Ancient Rome might look a little askance at official costume, but it had no such tradition against the display of other insignia of office. The consul had his fasces borne by lictors and magistrates their curule chairs; the augur carried his curved wand, the lituus; the senator had his ivory rod, and so on. Such symbols are the Western equivalent for the official robes of Greece and the Near East, where insignia were less common (e.g., the O.T. high-priest had vestments, but no equivalent to the pastoral staff). The general Christian acceptance in the fourth century of the Western principle of not using special liturgical robes makes it a little surprising that the other Western practice of the display of symbols of office instead was not accepted. But that the church was very slow in adopting such things is clear from the evidence.

Crosses. Constantine set the example of using the cross in insignia, both by mounting it upon the imperial diadem (and on the shields of his troops), and especially by his use of it on the labarum, the most important of the standards borne before the emperors. This he now made to consist of a gilded cross surmounted by the monogram of Christ, from the arms of which hung a banner of purple silk. He also set a gilt cross above the figure of a dragon on a pole which had formed the cavalry standard of Diocletian’s army.

The church, however, did not quickly adopt this carrying of a cross from the ceremonial of the court into that of the liturgy. The first we hear of crosses in a Christian procession is some seventy years later. Chrysostom at Constantinople organized torch-light processions to counter the street-propaganda of the Arians, and these carried silver crosses, to the arms of which were affixed burning candles. But it is clear that this was not a transference to the streets of something already practiced in the liturgy, but a novelty devised to attract attention, for the crosses were specially presented by the empress for the occasion. The carrying of ‘handcrosses’ (perhaps originally reliquaries) by dignitaries in church came in during the sixth century, and we hear of crosses carried in procession in Gaul during the fifth and sixth century. One was carried at the landing of Augustine of Canterbury in Thanet in a.d. 596, but here again it is possible to suspect an ad hoc device to attract attention rather than a piece of customary ceremonial.

In the Ordo Romanus Primus, which though it was compiled c. a.d. 800 seems to reflect the Papal ceremonial of the seventh-eighth century with considerable exactness, there is twice mention of a number of crosses carried behind the Pope, apparently not by clerics but by lay servants. It reminds one of the eagles and other standards carried by slaves behind the consul and other Roman magistrates. It is a piece of secular rather than religious pomp. But there is nothing in the Papal procession at this date corresponding to the later Western processional cross at the head of the procession, or to the special Papal cross. These both seem to owe their origin at Rome to a suggestion which that lover of ceremony for its own sake, the Frankish emperor Charlemagne, made to Pope Leo III in a.d. 800. When the Pope tactfully agreed with the happy idea of his distinguished visitor, he was at once presented with a magnificent jewelled cross for the purpose. This he ordered to be carried before him annually at the head of the procession of the ‘Greater Litanies’ on April 25th (not yet kept at Rome as S. Mark’s day). From the Papal procession the idea spread to the parish churches of Rome, which all acquired ‘stational crosses’ for use in procession during the ninth century. But the practice must have been well established at France long before Charlemagne brought about its adoption at Rome. Not only have we the occasional mentions of processional crosses by Gregory of Tours and other authors of the fifth and following centuries; but every parish church has already its own ‘stational cross’ for use in the Gallican ‘Litanies’ on the Rogation Days, in Angilbert’s Ordo at S. Riquier in Picardy c. a.d. 805.

The bearing of a special cross before archbishops everywhere within their own province appears to be a copying of this special Papal custom inaugurated by Leo III c. a.d. 800. It had already come into general use before the eleventh-twelfth century, when it caused continual troubles in England between the sees of Canterbury and York.

Altar Crosses, The placing of a cross actually upon the altar during the liturgy is often said to be derived from the use of the processional cross, the head being detached from the staff after the procession and stood before the celebrant during the eucharist. It does not seem that this was the origin of the altar cross, though it was the custom in some thirteenth-fourteenth century churches. The placing of anything whatever upon the altar except the bread and cup for the eucharist was entirely contrary to normal Christian feeling down to c. a.d. 800. In the ninth century, however, this was so far modified that out of service time the gospel book, the pyx with the reserved sacrament and reliquaries began to be admitted as ornaments placed upon the altar itself. But we still hear nothing of altar crosses. For centuries precious crosses had sometimes been hung above the altar, as had crowns, lamps and other ornaments; and standing crosses now began to be set up near it. But the first definite reference to an altar-cross of the modern type appears to be by Pope Innocent III (then still a cardinal) c. a.d. 1195, who tells us that at the solemn Papal liturgy a cross between two candlesticks is placed actually upon the altar. The custom spread gradually through the West during the thirteenth and fourteenth centuries, though it hardly became universal before the sixteenth. The Roman custom, however, during most of the middle ages was to remove these novel ornaments as soon as the liturgy was over, leaving the altar outside service-time as bare as it had always been in the past. This removal was still practised in many French churches down to the eighteenth century, and survives in a few Spanish churches to this day.

The Pastoral Staff. We have seen that Pope Celestine c. a.d. 425 regarded the use of a special staff by a bishop in the light of a reductio ad absurdum of superstition. Rome has so far proved faithful to his ideas that the Popes have never yet adopted the use of a pastoral staff. Pastoral staffs, however, did come into use elsewhere. They seem to be mentioned first in Spain in the early seventh century. They were then borne by Spanish abbots and abbesses as well as bishops, as symbols of office. From Spain they appear to have been adopted first by the Celtic and then by the Anglo-Saxon churches, and to have spread over the West outside Rome in the eighth-ninth centuries.

The Greek episcopal staff has a separate origin. It is derived, as its form indicates, from the crutch or leaning-stick employed by the Eastern monks as a support when standing through the long offices. Eastern bishops being recruited almost entirely from the monastic orders, they retained as bishops the staff to which they were already accustomed, merely giving it a more expensive and dignified form.

The Episcopal Ring. Signet rings were, of course, worn by bishops as by other Christians from early times. The first mention of a ring as being, like the pallium and staff, a special symbol of the episcopal office is in the twenty-eighth canon of the Spanish Council of Toledo in a.d. 633.

None of these symbols of office, however, appear to go back to the period of transition from a pagan to a Christian world in the fourth century. They developed only by degrees, in the seventh-ninth centuries, when deliberate imitation of the pagan rites of antiquity is, to say the least of it, very improbable. It remains, however, to notice one set of Christian insignia which do go back certainly to the later fourth century, and to point out their significance.

Fourth Century Insignia. A document called the Notitia Dignitatum Imperii Romania a sort of combination of Burke’s Peerage, Imperial Gazetteer and Directory of the Civil Service, reveals that c. a.d. 400 certain high officials had the privilege of being preceded on occasion like members of the imperial family, by attendants bearing lighted torches and incense. When entering their courts to dispense justice these officials added to these insignia their Liber Mandatorum or ‘Instrument of Instructions,’ a document which they received on taking up their office, setting forth the general line of policy which the reigning emperor intended them to follow. The particular copy of the Notitia which happens to have survived seems to have been drawn up at two different times. The portions dealing with the Western part of the empire reflect conditions c. a.d. 405; those dealing with the East seem to refer to a rather later period. But a number of scattered references in much earlier writers make it certain that the distinction of being preceded by incense and torches is something which goes back for some centuries before this in the case of Roman magistrates.1

This custom seems to have been adopted by Christian bishops in some places towards the end of the fourth century, at which time the state was placing upon them some of the duties of civic magistrates in their see-towns. But though these distinctions would thus seem to have originated much more from the secular than the strictly religious aspect of their position, a religious turn was given to it by the substitution of the gospel book as the Taw of Christ’ for the Liber Mandatorum of the secular official.

The first fairly certain reference to the episcopal use of these insignia appears to be in a poem by the Italian S. Paulinus of Nola just after a.d. 400; and there can be little doubt that its adoption, in some churches at least, both in the East and West, dates from about or rather before this time. It may be connected with the introduction of a solemn processional entry of the bishop and clergy at the beginning of the liturgy, which replaced the old greeting of the assembled church after an informal arrival, as the opening of the synaxis. We are rather in the dark as to when this procession was first introduced, except that at Rome the chant which accompanied it, the introit-psalm, is said to have been an innovation of Pope Celestine I (a.d. 422-432). The procession itself may be rather older than the practice of accompanying it by a chant; and taking into account the normal delay at Rome in the adoption of new liturgical practices, we might well suppose that the procession was at least twenty or thirty years older in some Western churches.

At all events, c. a.d. 400 and perhaps rather earlier, the bishop on entering and leaving the church began to be preceded by the torches, incense and book of a magistrate, a practice which had originally no particular Christian symbolism at all. An exact modern parallel is the preceding of Anglican dignitaries in procession by a beadle or verger carrying just such a ‘mace’ as precedes the Speaker of the House of Commons or a Mayor. At its beginning the use of these episcopal insignia had no more significance than that of the cathedral verger ‘pokering’ the canon in residence to read the second lesson. But when we first meet these processional lights before the bishop in the Roman rite they have already become seven in number (instead of the Praetorian Prefect’s four and the lesser magistrate’s two). It may be that here the seven golden candlesticks of the Apocalypse have come in to give a Christian turn to the old secular emblem. The bishop is the earthly representative of Christ, as the eucharist is the earthly manifestation of the heavenly worship, and the adaptation would easily suggest itself.

The use of the seven processional torches at the bishop’s liturgy spread widely through the West from the ninth century onwards, chiefly through an adaptation of the Ordo Romanus Primus made c. a.d. 800 which formed the basis of episcopal ceremonial in France for some centuries to come, and which was more or less widely adopted from there in England and Germany. Its only survival to-day other than in the Papal mass is in the special pontifical ceremonial of the archbishop of Lyons (which is not ‘Gallican’ in origin as has been too often supposed, but represents the ceremonial of the Papal rite as modified for adoption in the palace chapel of Charlemagne, which was introduced at Lyons by Bishop Leidrad, c. a.d. 810).

Whether the use of seven candles upon the altar by Western bishops when pontificating has any direct connection with the Pope’s seven processional torches (as has often been suggested) seems more than doubtful. When candles first appear upon the altar at the Pope’s eucharist they are not seven but two; and the seven altar candles when they do appear in the Papal mass do not replace the seven torches, but are an addition to them.

The use of two torches carried before the presbyter as celebrant of the eucharist seems to perpetuate the original form in which this honour was paid to bishops. It was probably an unreflecting continuance of custom when bishops finally ceased to be the normal celebrants of the eucharist for all their people at a single stational eucharist, and parish priests became their regular substitutes for particular districts. As the bishop’s delegate, no doubt, any celebrant seemed entitled to the same marks of honour, even though originally these particular insignia denoted rather the bishop’s personal importance as a civic leader than his sacerdotal character as celebrant of the eucharist.

Another symbol of the same kind which may have come into use in the late fourth or early fifth century is the umbella, a sort of flat ‘state umbrella’ carried over the heads of Byzantine magistrates and officials. It was also carried in front of the Byzantine emperor as a symbol of authority. In this fashion it seems to have been used by some of the Popes as a symbol of quasi-ducal authority in Rome after the ninth century, as it was by the doges of Venice and certain other Italian potentates in the early middle ages. It is doubtless from this that its use above the arms of the Cardinal Camerlengo of the Roman church during vacancies in the Holy See is derived, since the Chamberlain acts as emergency locum tenens of the temporalities of the see during the interregnum. But it never became a part of the Papal liturgical insignia nor a general symbol of the episcopal office. It had in fact no more religious significance than the state umbrella and fan now carried behind the Viceroy of India in public. But in a small number of ancient parish churches round about Aries in Provence — that stronghold of the old usages of Romania — the umbella is still carried over the head of the parish priest (but not, it is said, of anyone else) when he goes to the altar to sing mass on great feasts. I should be prepared to believe that this custom has come down by unbroken tradition from the last days of the empire in these cases, though I know of no evidence to prove it.

Here again, then, in the use of symbols and insignia, it seems quite impossible to bring home to the fourth century church any imitation of the pagan mysteries. The carrying and exhibition of symbolic objects in processions and liturgical rites was a notable feature of the mysteries in so far as they were public cults — and indeed of classical pagan worship generally. But what emerges from the evidence is that the Christian church made no ceremonial use of such things in the fourth century at all. The only possible exception is the Eastern offertory procession of the Great Entrance, first attested in its developed form by Theodore of Mopsuestia (Cf.p.282.) early in the fifth century. Those who wish to may lay emphasis on the general resemblance of this to a mystery rite, though I have failed to find any particular pagan rite to which it can be compared at all closely in detail. For my own part, given the Syrian custom attested by the Didascalia in the third century, of the deacons bringing the people’s offerings of bread and wine from the sacristy at this particular point of the rite, I think the ‘Great Entrance’ much more likely to be simply a ceremonialised form of this purely utilitarian bringing of the bread and wine to the table when they were required for the eucharist, than anything derived from the procession of the ‘dead Attis’ or such-like mystery cult functions.

Apart from this, the only portable symbols which were adopted anywhere before the end of the fourth century were the gospel book and the torches and incense carried before the bishop; and these were taken over from the civil ceremonial of the magistrate, not from the pagan cults, and had no religious significance. It is only centuries afterwards, when the pagan mysteries had long been forgotten, that the natural symbolic instinct produced the carrying of such objects as crosses and pastoral staffs in the Christian liturgy.

The episcopal insignia first introduced two things into Christian eucharistic worship, portable lights and censers, which play a considerable part in later ceremonial both in the East and West. But torches and candles have also been used in catholic worship in other ways which have not all this origin.

At Funerals. The lighting of torches at funerals was a mourning custom common to all Mediterranean religions, to which pre-christian Judaism had been no exception. The contemporary Acta of S. Cyprian’s martyrdom (a.d. 258) reveal that the pre-Nicene Christian church also made no difficulty about accepting this universal token of mourning. It describes how after a hasty temporary burial Cyprian’s body was subsequently removed by the Christians ‘with candles and torches.’ There was no change made about this after the peace of the church. Eusebius describes the candles burning on golden stands around the bier at the funeral of Constantine in a.d. 337 and S. Gregory of Nyssa describing his own sister’s funeral in a.d. 370 tells how deacons and subdeacons two abreast bearing lighted candles escorted the body in procession from the house. The custom was universal both in the East and the West, and continues so to this day.

Here (at last) is something in catholic custom which is certainly of pagan origin. Both the bier-lights (which have never died out at state funerals in post-Reformation England) and the Western chapelle ardente, and the candles held by the mourners at the Western requiem and the Eastern panikhida have all a common origin in very ancient pre-christian pagan observance. Mourning customs are always one of the most persistent elements of older practice through all changes of religion, chiefly because they depend on private observance by grief-stricken individuals much more than on official religious regulation; and no ecclesiastic is going to go out of his way to rebuke harmless conventions which may do a little to assuage sorrow at such a time. (So e.g. the modern West African Christians, both catholic and protestant, wear white at funerals in Ashanti, simply because a plain white ‘cloth’ in place of the normal brightly coloured native dress is the traditional mourning of Ashanti pagan custom.)

At the Gospel S. Jerome writing in A.D. 378 from Bethlehem says that ‘throughout all the churches of the East when the gospel is to be read lights are kindled ... not to dispel the darkness but to exhibit a token of joy ... and that under the symbol of corporeal light that light may be set forth of which we read in the Psalter, "Thy word is a lantern unto my feet and a light unto my paths"." This is one of those little symbolical actions like the lavabo with which, as we have said, the fourth century churches soon began to overlay the bare outline of the pre-Nicene rite, a process in which the Jerusalem church was the pioneer. In this case the context suggests that these lights were not so much part of the official ceremonial as kindled and held by the people. It is therefore probably more closely connected with that popular pagan custom of lighting lamps and candles both at home and in the sanctuary as a general sign of religious festivity, than with the later Christian ceremonial carrying of two candles by acolytes at the reading of the gospel. It had from time immemorial been a pagan usage to hang lighted lamps about the doorways of the house on days of religious festivity, about which more than one of the pre-Nicene fathers make scornful observations. Popular piety carried on the practice to celebrate Christian festivals, though it was discouraged by the church. But this popular use of lighted candles with their natural symbolism of cheerfulness and joy was too harmless to be rigidly excluded (c/., e.g. the use of candles on Christmas trees in Wesleyan chapels) and they make their way into various minor ceremonies of the liturgy towards the end of the fourth century. It is e.g. at this time that the presentation of a lighted candle to the neophyte after baptism (as well as the pre-Nicene white robe) begins to be introduced; and it is likely that this kindling of lights at the gospel in the East of which S. Jerome speaks is another quasi-liturgical observance of the same kind, introduced about the same time.

The more strictly official carrying of two lights at the gospel is first mentioned by S. Isidore of Seville early in the seventh century, but since he mentions that they were extinguished as soon as the gospel had been read, this may have a purely utilitarian origin, like the use of the prelate’s ‘hand-candle’ (scotula), the origin of which seems to be lost in antiquity. Anyone who has inspected ancient liturgical books, with their close writing and frequent contractions of spelling, will understand the need of a light near the book even in daylight for the public reading of the text. It is possible that once more Rome was somewhat behind other churches in the adoption of the lights at the gospel. The absence of them at the paschal vigil mass (on Holy Saturday) is probably a little piece of conservatism at this most archaic service in the whole year, reproducing the customary absence of ritual pomp at the singing of the gospel at Rome perhaps as late as the fifth or sixth century. It is only in the Ordo Romanus Primus that we first hear of two candles carried at the singing of the gospel in the Roman rite. Here it is clear from the whole setting and from what is done with them that they have a ceremonial, not an utilitarian, purpose. They precede the subdeacons with the censer (the book is not sensed as yet), but the gospel is sung from the top of the ambo (pulpit) steps while the lights remain below. The gospel book preceded by lights and incense has in fact come to be treated as symbolic of the Person of Christ proclaiming the gospel. Probably the lights which had been carried before the bishop for two or three centuries by now had introduced this new idea in connection with the book of the gospels.

Illumination. We have already seen (p. 87) that the ceremonial bringing in and blessing of a lamp was a customary part of the ritual at a chaburah meal such as the last supper. But this continued in Christian liturgical use only at the agape, not at the eucharist. It survived at the vigil also, and was introduced into the public service of the lucernarium from the practice of Christian domestic piety when public evening services began to be held in the later fourth century.

Nevertheless, illumination was, of course, sometimes needed for practical purposes at the early morning eucharists of the pre-Nicene church, and was provided in the ordinary way, as the candlesticks and lamps of the church of Cirta show (Cf.p.24.). But there was no ceremonial or symbolical use of lights whatever at the eucharist in the pre-Nicene church. After the peace of the church a number of fourth century authors speak incidentally of the great quantity of lights, both candles and lamps, sometimes employed in the churches at Vespers and the Night Office. We have already noticed the lavish scale on which Constantine provided for the lighting of S. Peter’s (Cf. p. 310). But though there is an advance here from mere utility to decoration, there is nothing corresponding to the later symbolic use of altar lights; though perpetually burning lamps at the martyrs’ tombs are found before the end of the fourth century. Curiously enough neither the precedent of the seven-branched lampstand of the O.T. Tabernacle nor that of the seven lamps burning before the throne of God in the Apocalypse seems to have exercised any marked influence before the beginning of the middle ages.

Candles on the Altar. For reasons already stated the standing of any object whatever on the altar was entirely contrary to the devotional conventions of the early church. Lamps and candelabra were hung above it, and standard candlesticks were stood around — sometimes six or eight of them. But the altar itself remained bare of such ornaments for almost the first thousand years of Christian history in the West, and perhaps to an even later date in the East. This feeling of the special sanctity of the altar began to break down in Gaul in the eighth century in certain respects, but it is not until the ninth century that we find candlesticks being stood upon it, and for some while they were not common even in great churches. There was one which was placed upon the altar in Winchester cathedral c. a.d. 1180, but apparently as a special little ceremony on Christmas day only, and this is the earliest English reference to such a practice that I know. This custom of one altar candle (moved around with the book at low mass) became fairly common in France in the thirteenth century, and was still not unknown in England as late as the fifteenth century. It is said to survive to this day at low mass in Carthusian monasteries.

It is not, however, until the very end of the twelfth century (c. A.D. 1195) that we first find candles upon the altar at Rome; and then they are two in number at the Pope’s ‘stational’ mass on the most solemn feasts. By A.D. 1254 the number on such occasions had risen to seven. Further than that it never went. The Papal custom of two candles on the altar was widely adopted in the early thirteenth century, and lasted without change in some of the great French and Spanish collegiate churches down to the eighteenth century.

It is by no means clear how the current notion that two candles was the specifically ‘English Use’ originated. The multiplication of altar candles was in fact rather characteristic of England and the North generally, once the custom of having them at all had come in. Thus e.g., at Chichester before the end of the thirteenth century the custom on feasts was to burn seven tall lights each of two pounds’ weight of wax upon the altar and eight more in trabe (on a shelf above the altar-screen — the fore-runner of the Renaissance ‘gradine’). At S. Augustine’s Canterbury there were two such trabes with a row of six candles on each, and apparently a third row of six actually upon the altar. At Exeter early in the fourteenth century there were still no candles on the altar itself, but a row of ten behind it. Ordinale Exon., ed. J. N. Dalton (H.B.S. 1909), ii., p. 540. At Lincoln there were five; at S. David’s cathedral there were fourteen; and so on. There appear in fact to be instances from mediaeval England of every number of altar candles from one to twenty, except seventeen and nineteen.

If we enquire the reason for the widespread increase in the number of altar candles during the thirteenth century, it is to be found, I think, in the change in the shape of the Western altar from the antique fashion of a cube some 3 ft. square to that of oblong altars 10, 12, or more feet long, in the new gothic churches. The increase in the number of candles comes in first in the great churches, which were mostly being rebuilt about then in the new style, only because the new shape of altar came in first in the great churches, which always tend to set fashions.

Such things have nothing to do with religion or its practice (or even with what is called ‘loyalty’), as the mediaeval churchmen were sensible enough to perceive. But the portentous behaviour of nineteenth century English bishops and lawyers, and the ‘fond things vainly invented’ by some ritualists, have succeeded in impressing it upon the mind of most modern Englishmen that they somehow closely concern the genius of Christianity. Such questions were formerly decided by custom, by aesthetics or by mere convenience, not by courts of law. To the mediaeval taste a row of candlesticks looked better than two on a long altar, and so they had a row — of three, four, five, six, seven, eight, nine, ten or whatever number their finances or fancy or just the fashion of the moment suggested; or they varied the number on different days according to the rank of the feast or the dignity of the celebrant. In Germany and Holland in the fifteenth century some churches took to having hundreds; in the same period in Sicily and Sardinia some churches preferred to retain only two; and nobody questioned their right to do as they liked in either case. The modern Anglican celebrant can have six candles upon his altar like some of the Avignon Popes in the fourteenth century, or seven like the Popes at the end of the thirteenth century, or two like the Popes at the end of the twelfth century, or even none at all like the Popes at the end of the eleventh century — and be happily conscious that historically he is being just as ‘Roman’ whichever he does. If he really wants to be ‘primitive’ in such matters, he must celebrate facing the people across the altar — like all the Popes in every century — and with no candles and no cross (and no vases of flowers or book-stand) — like all the Popes for the first thousand years. What preposterous nonsense it is to try to erect sacristy orthodoxies and even tests of theological allegiance out of these minute details of pious furnishing, that have varied endlessly throughout Christian history and have never meant anything in particular by all their changes!

Lights as Votive Offerings, The burning of votive candles as well as other lights (and incense) at the tombs of ‘heroes’ and before the statues of the gods was a general practice in Mediterranean paganism, and was not unknown in pre-christian Judaism at ‘the tombs of the prophets.’ The introduction of this form of popular devotion at the tombs of Christian martyrs even before the end of the pre-Nicene period seems to be witnessed to by a canon (34) of the Spanish Council of Elvira c. a.d. 300 forbidding it (though this interpretation of the canon is not quite certain). The Council’s prohibition certainly did not end the practice, even in Spain. A century later the Spaniard Vigilantius of Barcelona, exhibiting that impatience of folk religion which is at once the strength and the limitation of puritans in every age, made a violent attack on the general use of this practice in his own day by Christians at the tombs of the martyrs. To this S. Jerome made an equally intemperate reply, comparing those who observed it to the woman who poured ointment upon the Lord, and their critic to Judas Iscariot. More than one bishop made attempts to restrain the practice, but as such expressions of popular piety are usually wont to do, it proved stronger in the end than all ecclesiastical regulations. The lighting of lamps and candles at the tombs of the saints became a normal feature of all such Christian sanctuaries and places of pilgrimage from the fifth century onwards, if not from the end of the fourth.

Candles offered to Images. The cultus of relics of the saints concerned the honouring of the actual bodies of the martyrs or portions of them, something which has been and will be again at the last day an integral element in their personalities. A further step was taken when the same honours were paid to statues and pictures of the saints and of our Lord Himself. The fourth century church accepted the cultus of relics without much question, but it was much more reluctant to allow this second step to be taken, being still very sensitive on that question of idolatry’ upon which the conflict of the martyrs had turned. Pictures of our Lord and of the saints had been known as decorations (in the catacombs and elsewhere) and means of instruction (e.g. the baptistery at Dura) since the late second or early third century at the latest. As such, pictures and statues continued in use during the fourth century, though there were protests about this, and the Spanish Council of Elvira had forbidden such decorations in churches. But there is no single case, I think, of that ecclesiastical tolerance and even encouragement then given to the popular cultus of relics being extended to the cultus of pictures or statues of Christ or the saints during the fourth or the first half of the fifth century. There is, too, a noticeably academic tone about Christian homilies on ‘the peril of idolatry’ in this period, which contrasts with the urgency of clerical denunciations of abuses in connection with the relic cult, and suggests that any tendency towards an undue veneration of pictures and images was not a very widespread problem in the church, before the fifth century at all events. The distinction of Christian ideas and practice from those of a still living and observable paganism was as yet too obvious to need much emphasis. It was only after the disappearance of paganism that disputes began about the Christian use of images — a point which needs more consideration than it has received in most histories of the controversy.

There remained, however, in the new Christian world one particular survival from the past which was outside the control of the church, and which was bound sooner or later to raise in some form the whole question of the cultus of images. The emperor-cult had always been the center of the practical problem of ‘idolatry’ for Christians. The usual test for martyrs had been whether they would or would not ‘adore’ the emperor’s image with the customary offering of incense. But the Notitia Dignitatum (c. a.d. 405-425) reveals that this particular method of demonstrating loyalty had survived in full working right through the period of the conversion of the empire. In the fifth century the portrait of the reigning emperor was still set up in the courts of justice and in the municipal buildings of the cities surrounded by lighted candles, and incense was still burned before it. The Arian historian Philostorgius brings a charge of idolatry against the orthodox of Constantinople in his day (c. a.d. 425) in that they burn incense and candles before the statue of the emperor Constantine, the founder of the city. (It is worth remarking that this seems to be more than a century before we have any definite evidence of a similar cultus paid to specifically religious pictures and images.) One can see how this had come about. When Constantine and his successors became personally Christians, they still as emperors remained ‘divine’ (or at all events the working center of the old state religion) for that large majority of their subjects who still remained pagan. For these the old forms of reverence simply remained in use. To change them might have been politically dangerous; it would certainly have been unsettling to pagan public opinion. And now that the emperor publicly disbelieved in his own divinity, many Christians found it more possible to pay the conventional ‘adoration’ to the imperial portrait as a matter of etiquette.

Yet this cultus of the emperor’s ikon was by tradition a religious veneration and was well understood to be so. It was bound to suggest the lawfulness of a similar cultus to the ikons of the King of heaven and the saints, and we do in fact find it brought forward as an argument in favour of the cultus of Christian images, once that began to be debated. This is not the place to consider the immense disturbance which the facing of that question occasioned all over Christendom in the eighth and ninth centuries, or the rather different lines on which it was settled in the East and West respectively. All that concerns us now is the extent of the connection of such cultus of images with the official liturgy of the church and the date when it began.

In the West there is virtually never any such connection at all. The Western church has officially practiced and encouraged the cultus of images by the clergy and laity in a variety of ways; but it has always kept it dissociated from the eucharist and the office. At the most all that could be cited is the setting of a crucifix upon the altar during the celebration of the eucharist, and its incidental censing during the censing of the altar. But even this slight connection does not begin until the thirteenth century.

In the East the connection is stronger. Not only does the veneration of ikons play a much greater and more intimate part in the personal devotion of the Orthodox East among clergy and laity alike than is common in the West, but their censing and veneration in a carefully prescribed order is laid down as an official part of the orthodox liturgy, both at the eucharist and at the office, as well as at other services. They are regarded not as mere reminders of what they portray, but as actually mediating the participation of their originals in the earthly worship of the church. Accordingly their veneration is an integral part of divine worship, just as rejoicing in the fellowship of our Lady and all saints and angels will be a real part of the joy and worship of the redeemed in heaven, which the earthly worship of the church ‘manifests’ in time. But here again it is doubtful if this conjoining of the veneration of images with the official liturgy is really ancient in the Byzantine church. It probably began in the ninth century, as part of the great renewal of emphasis on the cultus of images which accompanied the final overthrow of the iconoclast emperors. In any case it can hardly be older than the introduction of the custom of a preliminary censing of the altar and sanctuary, which is first mentioned in Syria only in the late fifth century but probably was not adopted at Constantinople till the sixth-seventh century.

The use of incense both for domestic purposes and in the cultus goes back for some centuries before the Christian era all round the Mediterranean basin. In the Near East it is much older than in the West, doubtless because the materials — gums and spices — are indigenous to those countries and not to the West. Its religious use in the Old Testament need not detain us, since it has no early connection with its use in Christian worship other than through the use of Old Testament symbols in various ways by the writers of the New.

At the chaburah meal. There is, however, a domestic use of incense in Judaism which is worth recording because of its possible connection with the last supper. The burning of spices in the room after the evening meal was a common custom in all the Mediterranean countries, but among the jews it was — like everything else — given a religious colouring, especially at the domestic rite of supper on formal occasions, of the type under which the chaburah meeting was included. The ceremonial introduction and blessing of a lamp has already been spoken of. It was at this moment that the spices also were introduced and blessed and burned. In the first century a.d. the question was disputed between the rabbis as to the order in which the lamp and the spices (or the chafing dish in which they were burned) were to be blessed. The school of Shammai held that first the lamp was to be blessed, then the ‘Thanksgiving’ was to be said, then the spices were to be blessed and burned. The school of Hillel held that both lamp and spices were to be blessed and used before the ‘Thanksgiving’ was said. This was not an exceptional rite but one of such normal occurrence that the omission of the bringing in of spices (to save unnecessary labour on the Sabbath) at the Friday evening meal with which the Sabbath began, became a special sign of the Sabbath; as their reappearance at the Saturday evening meal was a sign that it was over. The reappearance of the burning spices on Saturday evening was especially associated with the habdalah, the prayer with which the domestic keeping of the Sabbath ended. In the form of the ‘habdalah spice-box’ this domestic use of incense has descended into the practice of the modern orthodox Jewish home, though it is not now burned, but only smelled at. The last supper was a formal chaburah meal, at which the ordinary rules for such occasions were observed, and it was not held on the eve of sabbath, which was specially marked by the omission of the burning of spices. It is true that the N.T. accounts do not mention this; but then neither do they mention the bringing in and blessing of the lamp, with which the spices were closely associated, though we may infer that there must have been one. They are not meant to be full reports of every detail of the meal.

In Christian Worship. It is probably due to familiarity with the hallowed usage of incense in the Temple worship and also in this domestic way at the chaburah meetings of the primitive church at Jerusalem, that there is no trace in the New Testament of hostility to the use of incense in worship. It is even taken for granted as playing a prominent part in the ideal Christian public worship of heaven. Such hostility developed later in the gentile churches during the persecutions. The mere fact that the ordinary test for a Christian was the command to burn incense to a heathen divinity was sufficient to cause it to be regarded with something like horror, despite the precedents of the Old and New Testament. These were allegorized away as referring only to ‘prayer,’ and the rationalistic arguments of pagan philosophers against the employment of incense in pagan worship were rather curiously seized upon as part of the Christian apologetic for its disuse. Turificati, ‘incense-burners,’ without further description, became a technical name for the apostates who by obedience to the magistrate’s command had forfeited not only the heavenly crown of martyrdom but all participation in the earthly worship of the church. Nothing can be more certain regarding the worship of the pre-Nicene church than that incense was not used at it in any way during the second and third centuries.

It was only after the peace of the church that the burning of perfumes in Christian churches began (Cf. p. 310). It must have become fairly widespread before the end of the fourth century for we hear of it almost simultaneously at Jerusalem and at Antioch in the East and at Milan and Nola in Italy. But there is nothing in most of these fourth century references to suggest more than a ‘fumigatory’ use of incense to perfume the churches. We do not even know that it was burned during service time, and not simply as a preparation for the assembly of a large and somewhat mixed gathering of people in a not too-well ventilated building. This is much more analogous to the domestic than the liturgical use of incense. The use of it borne before the bishop as a mark of honour, which comes in at about the same period, is nearer to a ritual usage, but even this is a borrowing from secular customs and not religious in its origin.

By the end of the fifth century some use of incense in Christian churches appears to have been more or less universal. But it is clear that in the large majority of cases this had still no more directly religious significance than, e.g., the use of music. It was now an accepted part of the general setting in which the eucharist was held; but the Old Testament notion of incense as in itself an offering to God (whether in combination with other sacrifices or alone) had hardly made its appearance. The text of Malachi i. 11 ‘in every place incense shall be offered unto My Name and a pure offering’ had, as we have seen, done yeoman service ever since the second century in expounding the sacrificial nature of the eucharist as the ‘pure offering’; but the reference to incense had invariably been ignored or allegorised away.

There is, however, one exception to this way of regarding the use of incense. Lietzmann has rightly drawn attention to a passage in the Carmina Nisibena of the East Syrian S. Ephraem composed in a.d. 363, which reveals that this thoroughly Jewish idea of the smoke of incense as in some sense an atonement or ‘covering’ for sin was already fully accepted in these predominantly Semitic churches. Addressing Abraham, the contemporary bishop of Nisibis, Ephraem says:

‘Thy fasts are a defense unto our land,

Thy prayer a shield unto our city;

Thy burning of incense is our propitiation;

Praised be God, Who has hallowed thine offering.’

Clearly this propitiatory ‘censing’ here is a liturgical function which the bishop performs on behalf of his flock, like prayer or the conseciation of the eucharist. A large number of other Syrian texts of the same character can be cited from the late fifth to the eighth century, all indicating the acceptance of the same idea of incense as a ‘sin offering.’ In this period the notion passed into the Christian liturgies. A ‘prayer of incense’ found in the oldest MS. (ninth century) of the Jerusalem Liturgy of S. James runs thus: ‘Thou that art made High-priest after the order of Melchizedek, O Lord our God, Who offerest and art offered and receivest the offerings; receive even from our hands this incense for a savour of sweetness and the remission of our sins and those of all Thy people.’ A variant of this idea is to be found in the Alexandrian Liturgy of S. Mark: ‘We offer incense before the face of Thy holy glory, O God; and do Thou accepting it upon Thy holy and heavenly and spiritual altar send down upon us in return the grace of Thy Holy Spirit.’ Other examples could be cited from all the Eastern rites.

Similar ideas reached the Gallican churches about the tenth century, probably from Eastern sources, and began to penetrate into the liturgies in the same sort of phrases. I cite the two following because these alone eventually passed from Gaul into the official Roman rite of the Pian missal in the sixteenth century, and so became more or less universal in the West. (a) A blessing of incense at the offertory: ‘By the intercession of blessed Michael the archangel standing at the right hand of the altar of incense and of all His elect, may the Lord graciously bless this incense and accept it for an odour of sweet savour. Through Christ our Lord.’ (b) During the censing of the oblations which follows: ‘May this incense which Thou hast blessed ascend up unto Thee, O Lord, and may Thy mercy descend upon us’; where the Egyptian idea of an ‘exchange’ of incense for grace seems to be latent though somewhat vaguely expressed.

In the development of the Christian use of incense we seem therefore to be able to trace the influence of three different factors: (1) The domestic or ‘fumigatory’ use. (2) The ‘honorific’ use of it before the bishop, which no doubt made it easier to transfer the idea of burning incense before the altar as a mark of reverence and so of an offering to God. There can be little doubt that this is the genesis of the Western censing of the altar. It is probable, too, that the contact with the instincts of folk-religion in the popular martyr-cult assisted in this. The custom of burning incense at a martyr’s tomb in his honour, which is attested in some places in the fifth century, shades off easily into the idea of an ‘offering’ to the saint to procure his intercession. (3) The purely Old Testament idea of incense as a sin-offering, which begins to infiltrate into Christian worship in Syria in the fourth century, and spreads gradually over the East and then penetrates into the West. Though this idea is accepted in isolated phrases in the liturgical texts, and has certainly — combined with (2) — operated to affect ceremonial in obvious ways both in the East and the West, it has never been formally accepted as a doctrine anywhere. It is noteworthy that in the conservative Roman rite all blessings of incense and censings of persons and objects were still unknown as late as the twelfth century, though by then they were more or less universal everywhere else. In the Papal mass of the twelfth century incense was still used as it had been everywhere (except in Syria) in the fifth century, only to scent the air and as a mark of honour carried before the bishop and the gospel book.

Such post-Reformation Anglican use of incense as there was before the later nineteenth century did not develop so exclusively as one might expect along the lines of the early ‘fumigatory’ use, though this was commonest. But the puritans under the Laudian regime were loud in their denunciations of censings ‘to’ altars, which suggests that the Carolines were influenced chiefly by Eastern precedents. It is a pity that we have no detailed description of the use of censing at Ely Cathedral, where it continued at least down to a.d. 1747. It ended because ‘Dr. Thos. Green, one of the Prebendaries and now (1779) Dean of Salisbury, a finical man, tho’ a very worthy one, and who is always taking snuff up his Nose, objected to it under Pretence that it made his Head ache.’

This brief and inadequate survey of the development of the accessories of ceremonial will have served its purpose if it makes clear how far it was from the intention of the fourth century church to convert men from heathenism by any imitation of the pagan ceremonies to which they were accustomed. The whole core and substance of the ceremonies as well as the rites of the eucharist in the fourth century were continued unchanged from pre-Nicene times; they can be traced back uninterruptedly through the formation of the ‘four-action shape’ of the eucharist to the chaburah rite of the last supper. Even such things as vestments, lights and incense in their use at the eucharist only begin to take on a properly ceremonial or symbolic character after the fifth century (at the very earliest), by the lapse of time through several generations. They have all either a utilitarian or secular origin in their liturgical use, and are given a particular Christian meaning only through the inveterate instinct of men to attach symbolic interpretations or at least a ceremonious performance to all public acts which are regularly repeated.

Yet there undoubtedly was a measure of assimilation both in practices and beliefs to the old pagan folk-religion during the fourth century. But it is in the practices of the martyr-cult, not in the eucharistic liturgy, that this is to be found. It is certain that in this field pagan practices and ideas did in the end succeed in naturalizing themselves within catholic Christianity, and came to be not only tolerated but encouraged by the clergy after the fifth century. The whole apparatus of the cultus of images, relics, holy wells, etc. in the forms which it was allowed to assume during the dark ages has a recognizable relationship to the same things in pre-christian paganism. But it is relevant to remark that just those elements in paganism which were taken over into Christian popular devotion were many thousands of years older than that ‘official’ paganism of the emperor-worship and the Olympian gods and the Eastern mysteries which the church overthrew. These popular practices had been assimilated by pagan ‘theology,’ as it were, and underlay it and survived it, just as they have survived conversion to Christianity, and also conversion to Judaism and Islam. Similar practices of offerings of lights and incense at the reputed tombs of welis and saints and prophets and marabouts are to be found in the popular mohammedanism and Judaism of the Near East and North Africa to this day.

It is not a sufficient defense of such practices in themselves to say that they are an instinctive popular way of practicing any religion, which has come down unchanged from the morning of the Mediterranean world. Yet this does make clear the process by which they passed over into Christian usage. It was not by way of the liturgy, which was under the control of the clergy, but through the individual expressions of piety of a multitude of half-instructed converts in the latter half of the fourth and especially the fifth century. The church allowed personal piety free play — how could she do other? — outside the liturgy; and in various ways it took the old instinctive lines. But these found their only point of contact with Christian public worship at the shrines of the martyrs. This is a rather different thing from the old charge of the deliberate paganising of Christian worship. It should always have been obvious to intelligent students of the period that when the clergy were preoccupied (as they were in the fourth and fifth centuries) with deeply philosophical problems of the nature and being of God and their relation to the incarnation, Plato and Aristotle were likely to present a much greater temptation to the fundamental paganising of Christian thought by the clergy than the lower strata of the old peasant superstitions which haunted the countrysides, but which had been despised by all educated pagans for centuries.

It would certainly have been more satisfactory to the modern mind if the church had taken a firmer line with these things in the fourth and fifth centuries, and prevented their recrudescence within Christianity; though to one who considers the actual field of their infiltration in the contemporary setting the practical difficulty of preventing it seems very great. The academic critic must make his reckoning with the fact that the actual compromise with them achieved in the fifth and following centuries is in itself no more, but also no less, defensible than the failure to deal firmly with the similar superstition that ‘An angel went down at a certain season into the pool of Bethesda and troubled the water: whosoever then first stepped in was made whole of whatsoever disease he had’ (John 5: 4). In the dark ages when ‘not many wise men after the flesh’ were available, the church was content to believe with the apostle that ‘God has chosen the foolish things of the world to confound the wise, and base things of the world and things which are despised hath God chosen, yea and things which are not’ (1 Cor. I:26 sq.). It may be a pity, but it is a fact, that it is impossible to reduce Christianity either to a spiritual philosophy or even to a pure theology. It is always a religion, which means partly a practice, for — amongst others — the immense numerical majority of uneducated people, who have their own place and office in the Body of Christ. What the church of the dark ages did not do, at all events in the West, was to allow such practices any foothold in the liturgy of the eucharist. Even in the East they remained on the margin of the liturgy.

The deliberate invention of symbolical gestures and actions and ceremonies in the liturgy to express and evoke adoration, purity of intention and so forth, is something which begins, as we have seen, in the fourth century with the transformation of the eucharist into a public worship. It is a subject with immense ramifications and fascinating bye-ways into which this is not the place to enter. But I think it can be laid down as an almost invariable rule that when each separate instance (e.g., genuflection, the lavabo, censing of the altar, etc.) is traced up to its beginnings, they have always the same history. They begin in Syria, usually in the fourth-sixth century, and radiate outwards, south to Egypt and north to Byzantium. In the West (to which they came sometimes by way of Byzantium, sometimes from Syria, and often first to Spain) the great Western center of interest in such devotional side-issues is always France, the first home or at least the chief propagator of so many modern popular devotions — the Rosary, the Sacred Heart, ‘Reparation,’ and so forth. From France they spread outwards to England, to Germany, to North Italy — and ultimately to Rome.

We shall not get very far in understanding the inner process of the history of the liturgy unless and until we understand that it expresses and must express something of the life of the Christian peoples; and that their natural characteristics do to a large extent enter into their religious life to be supernaturalised by grace. The perfervid devotionalism of the Syrian, which comes out so strongly, e.g., in Ignatius of Antioch c. A.D. 115 (and for that matter in Saul of Tarsus and some of the O.T. prophets) — the ceremoniousness of the Byzantine, with his love of etiquette — the naiveté of the Copt and his love of repetitions — the French mutability and love of some new thing — that special ‘tenderness’ of English devotion, which manifests itself in a love of rather sentimental hymns and vocal prayers in the first Anglo-Saxon private prayer books that we have — the prosaic practicality and the almost stuffy conservatism of the local church of Rome — these things do not change from century to century, and they are not annihilated when men come to pray. It is no accident that the deacon still leads the intercessions of the people in the Byzantine litanies with the very gestures and phrases prescribed by etiquette for the spokesman of a deputation to the Eastern emperor — that the Gallican ceremonial and rites are florid and have a greater number of variable prayers than any other — that the chivalrous doctrine that the Mother of God was never under the guilt of original sin appeared first in Anglo-Saxon England, where the treatment of women was much in advance of that common in Europe in the eleventh century — that Irish devotion has enthusiasm but practically no ‘liturgical sense’ whatever right through the centuries — that the Roman rite has about it still an archaic angularity and abruptness, a concentration on the performance of the eucharistic action rather than talking about it, which is no longer found in any other rite.

These matters of temperament are not only relevant to — they are the actual cause of — the course which the history of liturgical details has taken in Christendom. To ignore them is to make that history incomprehensible. But having understood their importance, we shall not be misled into making them a justification for misunderstanding the unity of the eucharist. They affect the details only of its performance. The main structure of the liturgy is always and everywhere the same, however much it be overlaid with local ways and decorations, because the eucharist is always identically the same action — ‘Do this’ — with the same meaning — ‘For the anamnesis of Me.’ In so far as the Christian Syrian and Byzantine and Copt and Englishman and Frenchman and Roman are all Christians and so partakers in the one eucharistic action and experience of the one Body of Christ, the Shape of the Liturgy by which that action is performed is bound to be the same in all essentials for them all.

WE have seen that the two halves of what we call the eucharistic rite were originally two distinct rites, the synaxis and the eucharist, either of which could be and frequently was celebrated without the other. They had different origins, served different purposes and were to some extent attended by different people. The eucharist, the Liturgy of the Body of Christ, was for the members of the Body alone. They had an absolute obligation to be present at it every Lord’s day, since the ‘vital act’ of the Body would be incomplete unless each member actively fulfilled in it what S. Paul calls ‘its own office,’ the ‘liturgy’ of its order. Those outside the Body, whether casual enquirers or enrolled catechumens, could attend only the synaxis and not all of that, since they were dismissed before the prayers with which it ended. Yet the synaxis is not rightly regarded either as a mere propaganda meeting for outsiders or even primarily an instruction service for the faithful, though the lections and sermons enabled it to serve both purposes. By intention though not in form it was an act of worship, the Liturgy of the Spirit, in which the church indwelt by the Spirit adored as well as proclaimed the divine redemption wrought through Jesus. The intercessory ‘prayers of the faithful’ which followed demonstrated, so to speak, the efficacy of that redemption by exercising His priestly power of intercession for all men bestowed upon the church, and on the church alone, ‘in Christ.’ Though the individual’s obligation to attend the Sunday synaxis may have been less strict than in the case of the eucharist, the faithful were expected to take part in this corporate witness to the fact of the Christian redemption. They were the only people qualified to exercise its consequence in the concluding intercessions, by appearing corporately before God, ‘accepted in the Beloved,’ to plead for the world.

We have traced out the exceedingly simple primitive structure of these two rites, which it may be convenient to set out again.

(When it was held separately the synaxis seems to have concluded with some sort of dismissal of the faithful.)

We have now to trace the addition to this primitive nucleus of a ‘second (23) stratum,’ as it were, of additional devotions, filling in, supplementing and in certain cases obscuring this bare primitive outline which concentrates so directly upon the essential action of the rite. In dealing with this ‘second stratum’ it is unfortunately much more difficult to avoid being technical. We have to take account of more facts, and the facts themselves are more complicated. The need of the period in which the ‘second stratum’ was added (from the fourth century to the eighth) was to adapt the old pre-Nicene tradition of Christian worship to its new ‘public’ conditions and function. But this need was felt by different churches with a different intensity and at different times. And the practical break-up of the Christian empire in the fifth century — it still continued as a theory, so mightily had the universal dominion of Rome impressed the imagination of the world — forced the local churches to meet the new needs to some extent in isolation, so that different schemes of additions appeared in different regions.

Before the fifth century her existence within or alliance with an effective universal state had enabled the church readily to put into practice her catholic ideal by the intercommunication of distant churches. When the old Roman world began to break up, the Christian world even in the practical breakdown of communications was still quite aware of its own unity; local churches were still quite willing and eager in most cases to borrow from elsewhere improvements and novelties in things liturgical. The result is that though the regional churches were in practice becoming sufficiently isolated to develop a considerable amount of variety in the new prayers of this ‘second stratum’, there was also a good deal of borrowing and cross-borrowing in various directions, due to occasional contacts, which complicates the individual history of the local rites a good deal. A new observance in the liturgy, e.g. of Milan in the fifth century, may be something evolved locally to meet a local need. Or it may equally well be something borrowed from Rome to the south, because of Rome’s prestige as the Apostolic See; or from Gaul to the north-west, because it is new and interesting; or something brought back from Jerusalem by returning pilgrims, full of ‘the way they do it5 in the Holy City of men’s holiest dreams and emotions in that age. All this needs careful disentangling if we are to make out the true history of rites, and above all the true reasons for changes, and their effects. And often enough the fragmentary evidence enables us to give only an approximate answer to questions we should like to ask about when and where such and such an observance, destined it may be to affect the development of eucharistic rites for centuries to come, first took shape and why.

It is impossible, therefore, to avoid a certain measure of complication in dealing with this ‘second stratum’ of prayers in the liturgy, though I have done my best to make it intelligible to the non-technical reader, because it is an essential part of the history of the eucharistic rites which Christians use to-day. But first it is necessary to say something of the process by which the two halves of the rite, originally distinct, came to be fused into a single continuous whole, for this process is the background of the addition of the prayers of the ‘second stratum’ to the old universal Shape of the Liturgy which had come down from pre-Nicene times to all churches alike.

Strictly speaking there was no conscious or deliberate process of fusion. As whole populations became nominally Christian, there ceased to be anybody not entitled and indeed obliged as a member of the faithful to be present at both rites. Confirmation was now received in infancy along with baptism as a matter of course by the children of Christian parents. In a Christian population the only people whose attendance at the eucharist could be prevented were the excommunicated — those who for conduct or belief incompatible with membership of Christ’s Body had been deprived of their rights and functions in the liturgical act of the Body. In Cyprian’s phrase, they had been ‘forbidden to offer,’ and by consequence to make their communion, for we must not forget that in primitive terminology those whom we call ‘the communicants’ are always called ‘the offerers’’ — offerentes, hoi prospherontes, not communicantes, koinonoi. The change of term to ‘communicants’ reflects an immense shift of emphasis in devotion. It goes along with a change in the status of the laity from participants in a corporate act with the celebrant to passive beneficiaries of and assistants at his act. These changes were not completed before the mediaeval period, and indeed constituted between them the essence of that mediaeval way of regarding the eucharist which has proved so unfortunate in different ways all over Christendom.

The roots of these changes go deep, right into the subsoil of the modern church. As far back as the fourth-fifth century the laity in general, especially in the East, were becoming infrequent communicants, out of a new devotional sentiment of fear and awe of the consecrated sacrament, of which we shall say a little more later. Thus, though they remained in name the offerentes or prospherontes, the faithful did in fact largely cease to offer their prosphorai of bread and wine, at all events with the old significance and as a normal weekly rule. The introduction of the devotional novelty of a special ‘holy loaf made by clerical hands as alone sufficiently holy for sacramental consecration further robbed the survival of the lay oblation of bread and wine (in so far as it did survive) of significance. From being the matter of sacrifice and the substance of self-oblation, the layman’s prosphora sinks to the sphere of the Eastern eulogia and the Western pain benit, mere tokens of a holy thing which the unhallowed layman ought not to receive. It is not surprising that the distinction between the faithful and the excommunicate became too difficult to enforce so far as mere presence at the eucharist was concerned (and nothing else but presence was now in question). The dismissal of penitents (i.e. those under discipline) vanished from most rites in the fifth-sixth century even in form, and was no more than an empty survival where it remained.

The deacons continued to proclaim the dismissal of the catechumens before the intercessory prayers as in the pre-Nicene church, but there were ceasing to be any catechumens to depart. By the seventh century this, too, had become a mere form. But where the prayers were kept up in some way at their primitive position after the sermon, the deacon’s dismissal of the catechumens was generally maintained as a sort of prologue to them, though the bishop’s departure-blessing of the catechumens which preceded it usually fell into disuse. Where the precedent — set at Jerusalem as early as c. a.d. 335 — of transferring the intercessions from the synaxis to the second half of the eucharistic prayer had been followed, the deacon’s dismissal of the catechumens was apt to disappear altogether from the rite, as e.g. in the Syriac S. James.

With the disappearance or toning down of the dismissals the most emphatic mark of division between fully ‘public’ and specifically ‘Christian’ worship was weakened, and the two services held one after the other on Sunday mornings soon came to be thought of as a single whole, because the same congregation now attended the whole of both rites as a matter of course. This stage had been reached in many places by the end of the fifth century. By the end of the sixth the holding of either rite without the other had come to be regarded as an anomaly.

But in the fourth century this fusion was hardly begun. The distinction is fully recognized, for instance, by Etheria in her account of Sunday morning worship at Jerusalem in a.d. 385: ‘At daybreak, because it is the Lord’s day, all proceed to the great church which Constantine built at Golgotha behind (the site of) the Crucifixion, and all things are done according to the custom everywhere (at the synaxis) on Sunday; except that (here) the custom is that of all the presbyters who sit (in the stalls round the apse) as many preach as wish, and after them all the bishop preaches. They always have these (many) sermons on Sundays, that the people may always be well taught in the scriptures and the love of God. And the preaching causes a long delay in the dismissal of the ecclesia, whereby it is not given before ten o’clock or sometimes eleven. But when the dismissal has been done, in the way it is done everywhere, the monks escort the bishop to (the church of) the Resurrection (on the other side of the great paved court enclosing Golgotha) and when the bishop arrives to the singing of hymns, all the doors of the basilica of the Resurrection are thrown open. All the people go in, but only the faithful not the catechumens. And when the people are in, the bishop enters and goes at once inside the screens of the martyrium in the cave (of the Holy Sepulchre, where the altar stands). First thanks are given to God (i.e. the eucharistic prayer is said) and then prayer is made for all (i.e. the intercessions). Afterwards the deacon proclaims aloud. And then the bishop blesses them standing within the screens and afterwards goes out. And as the bishop proceeds out all come forward to kiss his hand. And so it is that the dismissal from the eucharist is delayed nearly to eleven or twelve o’clock.’

Here the two rites are not only distinct, but held in different churches. The synaxis is public, and at Jerusalem exceptionally lengthy. The eucharist is still exclusively for the faithful and comparatively short — less than an hour. Etheria is trying to be discreet in describing its details, in deference to the old discipline of not publicly revealing the content of the rite. But she manages to let her sisters in Spain know how ‘the way they do it in Jerusalem’ differs from things at home in Galicia — by the (rather overwhelming?) number of sermons, and the bishop’s processional exit from the synaxis and entrance at the eucharist, and by the postponement of the intercessions from the synaxis to the eucharist, the hidden consecration, and the final blessing.

The postponement of the intercessions to the eucharist had a practical advantage at Jerusalem, arising out of the local custom of transferring the congregation from one church to the other between the two rites. The catechumens could be left outside in the courtyard without the delay of getting them out of the midst of the synaxis-congregation before beginning the intercessions. Perhaps the transference of the intercessory prayers to the eucharist began at Jerusalem out of this utilitarian motive. Shorn of the prayers the synaxis became a wholly ‘public’ service, and all the strictly Christian worship was concentrated in the eucharist.

But when this local Jerusalem custom began to be imitated in other places where there was no second church, and both services were held one after the other in the same building, the transference of part of the synaxis into the second half of the eucharistic prayer must have gone some way of itself towards fusing the synaxis and eucharist, by eliminating precisely that point of the rite at which the distinction between the ‘public’ and Christian worship had hitherto been made.

Yet whatever other factors may have helped to break down the distinction between synaxis and eucharist, it was undoubtedly the disappearance of adult catechumens which finally ended the need for any such distinction. The moment at which the whole population (to all intents) could be said to be nominally Christian naturally varied a good deal in different places. A fair test is the lapse into disuse of the ‘discipline of the secret’ — the old rule of never describing the eucharist openly in the presence of the unconfirmed or in writings they might see. At Rome this stage had been reached c. a.d. 450. The sermons of Pope S. Leo preached in the presence of any catechumens there might be, and also his official correspondence, speak of the details and doctrine of the eucharist with a complete absence of that mystification still indulged in by S. Augustine in his sermons and by Pope Innocent I in his letter to Decentius only a generation before. Yet at the opposite end of Christendom a few years later Narsai makes it clear that at Edessa there were still adult catechumens, and their expulsion from the liturgy was still a living reality. And the eucharist could still be celebrated there without being preceded by the synaxis, at all events at the paschal vigil, just as we find it at Rome in the second and third century. But even at Edessa after another 200 years there are no longer any catechumens, and the deacons are ceasing in practice to command their withdrawal from the liturgy, though it still remains in the text of the Edessene rite.

The fusion of synaxis and eucharist was thus taking place gradually c. a.d. 400-500 in most places. But during this period each continued to be celebrated without the other on occasion. Thus the Byzantine historian Socrates (c. a.d. 440) says that the synaxis ‘without the mysteries’ is still held every Wednesday and Friday (the old ‘station days’) at Alexandria in his time. But he looks on this as an old local peculiarity and seems to have no idea of its former universality. Perhaps he was confused because in the contemporary Byzantine rite such a synaxis without the eucharist already always took the form of the ‘Liturgy of the Pre-sanctified,’ i.e. of a synaxis followed by communion from the reserved sacrament, as the close of a fast day — which rather disguises the nature of the rite.

In the West the synaxis apart from the eucharist persisted chiefly in Lent, as a relic of the old instruction-classes for the catechumens. It was ultimately restricted first to Holy Week and finally to Good Friday only. When the Roman church first began to observe Good Friday as a commemoration of the Passion separate from that of the Resurrection on Easter Day (instead of both together at the paschal vigil) — in S. Leo’s time c. a.d. 450 this change had not yet been made — the old Roman texts of the paschal vigil were transferred bodily with a minimum of adaptation to a synaxis without the eucharist on the Friday, to make way for the new series of lections at the Saturday vigil drawn up in the church of Jerusalem in S. Cyril’s time (Cf. p. 339). In the sixth century this synaxis composed of the old Roman texts of the second century for the Saturday vigil continued to be the only strictly official observance on Good Friday at Rome. This synaxis ended with the intercessory ‘prayers of the faithful’ in the Papal rite. But the communion of the Pre-sanctified had attached itself to the synaxis in the parish churches of Rome on Good Friday long before it was accepted in the official rite of the Pope. It was probably a survival, unchanged in the popular tradition of devotion since pre-Nicene times, of communion at home from the reserved sacrament on those fast days on which there was no celebration of the eucharist, which had transferred itself to the parish churches when domestic reservation began to be given up (? in the fifth century).

So much for the synaxis without the eucharist. The eucharist without the synaxis seems to have disappeared everywhere in the East after c. a.d. 500. It lasted longer in some places in the West, but only as a special survival on Maundy Thursday. On that day in some Western churches there were three eucharists — one for the reconciliation of penitents in the morning, one for the consecration of the chrism at mid-day, and one in commemoration of the last supper in the evening. At the first eucharist there was no synaxis, the long rite for the reconciliation of penitents taking its place. At the second the synaxis precedes the eucharist in the normal way. At the third the eucharist is celebrated without the synaxis, beginning, as we should say, at the offertory. This is, of course, the ‘typical’ eucharist of the year, and its holding in this primitive fashion may have been due to a lingering tradition of what constituted the rite of the eucharist proper. Since the synaxis had already been held that day, there seemed to be no need to impose it again on the congregation and celebrant already weary with the long fast. But we hear no more of this evening eucharist on Maundy Thursday after the ninth century. Apart from the single exception of the Liturgy of the Pre-sanctified (really a synaxis without the eucharist, though the appended communion from the reserved sacrament partly disguises the fact) the two rites had finally become a single indivisible whole all over Christendom well before a.d. 800.

There is one further aspect of this fusion of the synaxis and eucharist which should be mentioned, though I am not in a position to answer the further questions which it raises. The period during which this fusion came about is precisely that in which mere presence at the eucharist, instead of the old liturgical and communicating participation in the eucharistic action, definitely established itself in most places as the substance of the ordinary layman’s eucharistic devotion. Was there some connection between the two movements? At the pre-Nicene synaxis a passive part was all that was possible for the congregation; the reader, the singer of the gradual, the preacher, necessarily acted while the rest listened. It was only when the intercessions were reached that even the pre-Nicene synaxis became an effectively corporate act.

The transference of these intercessions into the second half of the eucharistic prayer, which was essentially the celebrant’s own individual contribution to the corporate act, certainly went far to destroy their corporate nature. In the liturgy of S. James and even in S. Cyril’s account of the matter they have become simply a monologue by the celebrant, in which the people have nothing to do but listen. I cannot think it is entirely accidental that this impulse towards ‘non-communicating attendance’ should apparently have begun in Syria, and that in the same period the Jerusalem rite, soon to be so widely imitated in the East, should have undergone this particular change. For it cannot have been without some effect that this most influential liturgy should have substituted at the very point at which the older rites prayed for the communicants and the effects of a good communion a very lengthy intercession for all sorts of other concerns, based on the novel doctrine of the special efficacy of prayer in the presence of the consecrated sacrament. This idea was taken up by the preachers, e.g. Chrysostom; and it has received ‘extra-liturgical’ developments in the mediaeval and modern Latin churches. It is not possible to deny its devotional effectiveness, though it may not be so easy to justify on theological as on psychological grounds. That is not here our concern. The point is that in the fourth century it was new, and the expression of it by substituting another idea for that of communion just before the communion act itself was new too. Coming at that particular point it can hardly have been without some effect on eucharistic devotion among those who paid attention to the prayers. Doubtless there were all sorts of psychological influences at work in producing the new idea of the laity’s wholly passive function at the eucharist — the instinctive feeling that communion was not for everybody, the new language of ‘fear’ of the sacrament, and so on — as well as a certain inevitable lowering of the temperature of devotion in an established’ church which was coming to include the average man as well the naturally devout. But it may very well be that amongst those influxes we have to reckon with some unforeseen effects of the liturgical changes made in the structure of the Jerusalem rite. And it remains a fact, explain it how we may, that the passive receptiveness — the being reduced to mere listening — which was always necessarily the layman’s role in the first part of the synaxis, became his role also the-eucharist proper (which it had never been before) just in the period in which synaxis and eucharist began to be regarded as parts of a single rite.

The tradition of the liturgy was as tenacious of its inherited forms in the fifth century as it had always been, and so the process of adaptation to new needs took the form of additions to the old nucleus much more often than of substitutions for it. This is noticeable in the synaxis. As the pre-Nicene church had transmitted it from the synagogues of the apostolic age, this rite might well seem unnecessarily abrupt in its opening. And however faithful to its origins and well adapted to its pre-Nicene purpose, it was everywhere defective in the elements of vocal praise and prayer, especially where the intercessions with which it should have concluded had been transferred to the second half of the eucharistic prayer. Once the decline of the catechumenate began to make it unnecessary to continue the old restrictions on these aspects of worship at the synaxis, it was right that attempts should be made to remedy these deficiencies. This was done by adding an ‘Introduction’ to the old nucleus, of a more directly worshipful character than the old conditions had allowed.

The uniformity of the ancient material in all churches will not have prepared the reader for the apparent complexity and diversity of the material added by the ‘second stratum’ in the various churches, as shewn in the table opposite p. 442 (which has been simplified in some columns by the omission of what are known to be mediaeval insertions). The items A, B, C, D, E, are found in all. These are the ancient nucleus. But of the other items prefixed to this, though each appears in several rites, none but the ‘hymn’ appears in all, and this in two different forms (β and 3) while some are in different positions in the rites in which they are founds e.g. the ‘prayer’ (§) in the Egyptian and Western groups.

Yet a few minutes’ study of this table with due regard to the approximate date of the appearance of the items in the various rites reveals that all this complexity has a comparatively simple explanation. It arises from the fusion in various ways of three different forms of Introduction. These three forms I have called I, II and III, distinguishing their contents by three different prefixed symbols to assist their identification. Those of I are α and β; the single item of II is marked §; and the components of III are I, 2, 3. All three forms (I, II, III) arose in the East during the fourth-fifth century, and have originally a geographical basis. We shall discuss them individually, and then ascertain the uses made of them in the Western rites, which form a fourth group (IV).

The first scheme consists of (a) a preliminary censing by the bishop or celebrant, followed by (β) the singing of a group of psalms, prefixed to the lections. Geographically it begins in what is for us the ‘far east’ of classical Christendom, though (α) the censing was afterwards adopted by the central group of Greek churches. We think and speak of these Greek churches as ‘the Eastern churches,’ but the Mesopotamians and other ‘far Easterns’ habitually called them ‘the Western churches,’ and the Greek theologians ‘the Western doctors.’ The Greeks always stood for Western and European ideas in the mind of these Semitic Christianities, for whom the Latin West was generally too remote to be taken into account.

We have already noted the special importance attributed to the bishop’s ‘censing’ by S. Ephraem in East Syria before a.d. 360 (Cf. p. 428). The same notion of censing as a propitiation and a preliminary even to private prayer is found in Syria in the fifth and sixth centuries. Thus in a.d. 521 the hermit Zosimas in Phoenicia ‘At the very moment of the earthquake at Antioch suddenly became troubled. .. called for a censer and having censed the whole place where they stood, throws himself on the ground propitiating God with supplications’ and afterwards told his companions of what was then happening at Antioch — an instance of his well-authenticated ‘second sight.’

But the first description of censing as a preliminary to the liturgy, and of the Oriental introduction to the synaxis as a whole, is found in those remarkable writings which succeeded in imposing a system of neoplatonic pagan mysticism upon all Christendom under cover of the name of Dionysius the Areopagite, the convert of S. Paul (Acts 17:34). The date and country, though not the identity, of this enterprising forger have been determined within narrow limits. He wrote c. a.d. 485 in that interesting strip of country behind and to the north of Antioch, which forms a borderland between N. Syria and Mesopotamia on the one hand and Asia Minor proper on the other, a region which has given to mankind not only such minds as Poseidonius and S. Paul and Nestorius, but a multitude of ideas and inventions. In his high-flown way Pseudo-Denys describes the opening of the synaxis thus: ‘The hierarch having ended a sacred prayer [? privately] before the divine altar, begins by censing there and goes throughout the whole enclosure of the sacred edifice. And returned once more to the holy altar, he begins the sacred melody of the psalms, the whole well-ordered ecclesiastical array chanting along with him the holy psalmic song.’

These psalms (β) survive in the marmitha psalms of the present E. Syrian rite, and also, apparently, in the ‘psalm of the day’ and saghmos jashou (‘dinner-time psalms5) of the Armenian rite (Ibid., p. 425). This rite, however, has at some time incorporated into itself the whole Greek scheme of the introduction (III), part of it being interpolated into the middle of the (β) psalmody of the original Armenian Oriental scheme (I). This (β) psalmody failed to establish itself in the Greek rites when (a) the censing was taken over by them, because the purpose of the psalmody was already served in the Greek liturgies by the ‘hymn’ (3) introduced at the same point in the Greek scheme c. a.d. 440. We find a preliminary censing before the liturgy mentioned at Constantinople c. a.d. 565, and it is likely to have been used in the Greek Syrian rites before then, since the idea of censing as a preliminary to prayer was well known there before that date. But at Constantinople the preliminary censing was performed by the deacon and not by the bishop as in the East, because the Greek bishop continued to enter the church only during the ‘entrance chant’ (I) of the original Greek introduction scheme (III), to which the censing (α) drawn from the Eastern scheme (I) had been prefixed. In Egypt we have a mention of the same sort of preliminary censing (α) before the synaxis in a document which in its present form can hardly be as old as the fifth century and seems more likely to be of the later than the earlier part of the sixth. Roughly speaking, the preliminary censing (α) of the Tar Eastern’ Introduction (I) had been incorporated into all Eastern rites but one before a.d. 600. Curiously enough the E. Syrian rite of Edessa has no censing (α), though it has the psalmody (β). It may be that it once existed in this rite, but the prefixing of a long formal preparation of the elements and the celebrant on the Byzantine model in later times has eliminated it. Or it may be that the (β) psalmody before the lections was introduced in the earlier fourth century before there was any use of incense in church; and that Addai and Mari thus preserves the first stage (β psalmody alone) of the type of Introduction of which Ps.-Denys in the next century gives us the developed form (α censing followed by β psalmody).

The first censing of the altar in the Western rites does not appear before about the tenth century in Gaul, and was not adopted at Rome until the twelfth; it spread not very rapidly in the derived Western rites during the middle ages. In view of the fact that some early Western ceremonials, e.g. at Milan, give this initial censing to deacons or minor ministers and not to the celebrant, it is conceivable that it began in the West as an imitation of Byzantine ways. But the Gallican ceremonialists were quite capable of developing this rite for themselves out of the old fourth century Western custom of merely carrying a smoking censer before the bishop in the entrance-procession as a mark of honour. And it is to be noted that the Western rites, unlike the Easterns, all kept the entrance-chant as the effective opening of the rite, and did not prefix the censing to it, as at Byzantium. The Western initial censing, late in making its appearance, never became more than an accompaniment to something else in the rite, a piece of ceremonial performed while something else was going on, and did not develop, as in the Eastern rites, into an item in the structure of the rite on its own account.

The second scheme seems to be locally Egyptian in origin. It consisted simply of the old pre-Nicene greeting (A), followed by a prayer (§), prefixed to the lections. In the earliest document of the Egyptian rite available (Sarapion, c. A.D. 340), we find that the synaxis begins with a prayer headed ‘First Prayer of the Lord’s (day).’ It runs thus:

‘We beseech Thee, Father of the Only-begotten, Lord of the universe, Artificer of creation, Maker of the things that have been made; we stretch forth clean hands and unfold our thoughts unto Thee, O Lord. We pray Thee, have compassion, spare, benefit, improve, increase us in virtue and faith and knowledge. Visit us, O Lord: to Thee we display our own weaknesses. Be propitious and pity us all together. Have pity, benefit this people. Make it gentle and sober-minded and clean; and send angelic powers that all this Thy people may be holy and reverend. I beseech Thee send "holy Spirit" into our mind and give us grace to learn the divine scriptures from (the) Holy Spirit, and to interpret cleanly and worthily, that all the laity here present may be helped; through Thy Only-begotten Jesus Christ in (the) Holy Spirit, through Whom to Thee be glory and might both now and to all the ages of the ages. Amen.’

This prayer immediately preceded the lections. It is, by its position, the earliest ‘collect’ we possess, and a surprisingly early case of disregard for the rule that prayer might not be offered in the presence of the catechumens. Perhaps Sarapion would have argued that this was not so much a prayer ‘with’ them as ‘for’ them, which was allowed. This Egyptian collect is not, as with our eucharistic collects, a variable prayer connected with the day in the ecclesiastical calendar, but one always the same, closely connected by contents and position with the reading of the lections which it introduces. The bishop prays not only in the name of his church — ‘we beseech Thee’ — but in his own name — ‘I beseech Thee’ — when he prays for himself, for the special gift of the Holy Spirit to interpret the message of the scriptures for the laity. The whole construction suggests an originally private devotion of the bishop which has been turned into a public and audible preliminary to the lections.

Sarapion gives us only the prayers said by the bishop-celebrant, not their setting in his ‘dialogues’ with the people and the responses and other parts of the corporate worship offered by the deacon and others. For this we must turn to the Greek Liturgy of S. Mark, the mediaeval descendant of the fourth century rite of Alexandria. In this late form the Introduction has been Byzantinised. The Byzantine formal entrance of the bishop has been introduced, accompanied by the sixth century Byzantine processional chant, the Monogenes. After this follows at once the original Alexandrian opening of the synaxis. The deacon cries ‘Stand up for prayer’ — calling the church to order, as it were; and the celebrant greets the church, ‘Peace be to all,’ and is answered, ‘And with thy spirit.’ The deacon repeats ‘Stand up for prayer,’ to which the people answer ‘Lord have mercy.’ Then the celebrant chants his collect:

‘Master, Lord Jesus Christ, the co-eternal Word of the everlasting Father, Who didst become like unto us in all things, sin excepted, for the salvation of our race; Who didst send forth Thy holy disciples and apostles to proclaim and teach the gospel of Thy kingdom, and to heal every disease and sickness among Thy people: Do Thou now also, Master, send forth Thy light and Thy truth, and illuminate the eyes of our understanding for the comprehending of Thy holy oracles, and enable us to hear them so that we be not hearers only but doers also of the word, that we may be fruitful and bring forth good fruit thirty and sixty and an hundredfold, and so be worthy of the heavenly kingdom . . .’

This is addressed to the Son, and contains a number of technical anti-Arian terms (e.g. synaidios) which were specially emphasised at Alexandria in the time of the teacher Didymus the Blind, in a particular phase of the Arian controversy c. A.D. 370. They suggest that this also is a fourth century composition, though somewhat later than Sarapion’s. Here again the Egyptian collect is directly connected with the lections, and asks for the fruitful hearing of the apostolic proclamation of redemption by the lessons and the sermon. This preliminary prayer (§) forms the whole of the Egyptian Introduction to the synaxis (II).

We shall deal later with its borrowing by the Western rites. It was never incorporated into the other Eastern Introductions. Instead the Egyptian rites themselves later incorporated the censing (a) from the Far Eastern scheme (I) and the entrance chant (1) and hymn (3) of the Greek scheme (III).

This is rather less homogeneous than the other two — or rather, perhaps, its full development was reached somewhat later. It consists of (1) a solemn processional entry of the bishop and clergy to the singing of a chant of some kind (Eisodikon) followed at once by the old opening greeting (A). There follow (2) a litany and (3) a hymn before the lections (B).

1. The Entrance Procession and Chant. We have seen that at Jerusalem in Etheria’s time the bishop’s entrance into the church of the Resurrection for the eucharist was specially delayed until all the people had taken their places, in order that he might enter in procession through their midst; and this though they had all been gathered together just before at the synaxis in the other church across the courtyard. Etheria does not describe the entrance of the bishop for the synaxis. But since she tells of a similar formal entrance of the bishop for two of the daily offices, and a processional departure of the bishop from both synaxis and eucharist it seems a fair inference that the synaxis also began with a processional entrance of the clergy — a typically Cyrilline touch of ceremony.

There is no procession or opening greeting in the liturgy of Ap. Const. (Bk. ii. or Bk. viii.), which begins straight away with the first lection, like the Roman synaxis on Good Friday. But S. John Chrysostom in homilies preached at Antioch c. a.d. 390 and at Constantinople soon after a.d. 400 refers to some sort of formal entrance and the immediately following greeting: ‘When the father enters he does not mount up to this throne before beseeching for you all this peace.’

The fifty-sixth canon of the Council of Laodicea in Asia Minor (c. a.d. 363) lays it down that ‘Presbyters ought not to enter and sit down on the bema (in their stalls round the apse) before the entrance of the bishop, but to enter with the bishop’ — an indication that the old informality was beginning to give way to the more dignified arrangements of a fully public worship in this region during the latter half of the fourth century. I can find no mention of any sort of formal entrance of the clergy for the liturgy in the writings of the Cappadocian fathers from the Eastern part of Asia Minor in this period. But S. Basil specifically tells us that much in the performance of the liturgy in his church of Neo-Caesarea — the chief church of this region — was rather ‘slovenly, owing to its old-fashioned arrangement,’ and this may be a point he has in mind. (This equation of ‘slovenliness’ with ‘old-fashioned’ is a permanent feature of the history of liturgy and is worth pondering by the ‘up-to-date’ of all periods — perhaps with some searchings of heart.)

In all this fourth century Eastern evidence, however, though the entrance procession of the bishop and clergy seems to be taking shape, there is no direct mention of a chant. The first talk of this seems to come from Rome in the time of Celestine I (a.d. 422-432). It may be that Rome for once set a new fashion in the liturgy. Yet we must remember that the processional entrance itself is attested in the East, e.g. at Laodicea, some sixty years before this, and that Etheria’s ‘hymns’ during the bishop of Jerusalem’s procession between the two churches may have continued while he passed between the ranks of people in the basilica of the Resurrection, though she does not say so. On the other hand the ‘silence’ of the offertory procession is emphasised by Theodore of Mopsuestia (Cf. p. 283) in a way which suggests that silent processions may have been found particularly impressive by Easterns at this time, though to modern Western eyes they usually seem slightly depressing.

We reach firmer ground as to the Greek entrance-chant in a.d. 535-6. The emperor Justinian at the close of his pro-Monophysite period, at a time when the monophysite patriarch Severus of Antioch was actually staying as his guest in the palace, composed a ‘prose hymn’ generally known from its first word, Monogenes. This class of composition is known to us in the West chiefly by specimens of Eastern origin (e.g. Gloria in excelsis at the eucharist).

Justinian’s hymn forthwith became the entrance-chant at Constantinople and Antioch. But shortly after this he changed sides, and proceeded to persecute the Monophysites with his usual cold-blooded efficiency. The Syrians and Egyptians soon came to execrate him as the incarnation of Byzantinism, and accordingly the monophysite rites of Syria and Egypt do not contain his hymn. To the ‘royalist’ Greek churches of Antioch and Alexandria his authorship was on the contrary a recommendation, and the Monogenes remained the first item of the Greek Introduction in the Greek rites. The Armenians (who had largely escaped Justinian’s missionary methods and therefore felt less strongly about his authorship) adopted it when they incorporated the Greek Introduction (III) into their own rite, though they have spasmodically patronized a monophysite or anti-Byzantine interpretation of the Creed.

2. The Litany. The origin of litanies and their first position in the rite, at the ‘prayers of the faithful’ after the sermon, are more conveniently dealt with later. Here we are concerned with the insertion of a litany in the Introduction to the synaxis between the entrance-chant and the hymn.

It will be noted that the Byzantine rite contains no litany at this point, and I know of no evidence that it ever did so. The Greek S. James, the rite of Antioch-Jerusalem, does contain one. There seems to be no evidence as to when it appeared in the local rite of Antioch, but it cannot be traced before the ninth century. Yet besides the fact that S. James now contains a litany at this point, which despite its Byzantinised text was not taken over in this position from Byzantium, there is the fact that when the Roman and Milanese rites came to take over the Greek scheme of Introduction in the fifth-sixth century they both inserted at this point a litany, whose text in each case is based on a Greek original. The Three Great Prayers,’ the Egyptian equivalent of a litany, occur at this point in the Coptic S. Mark. All these facts would be adequately explained by the supposition that the litany here originated in the local use of Jerusalem (as to which unfortunately we have very little evidence in the fifth-eighth century) and that it spread north to Antioch and (after a fashion) south to Egypt and West to Rome, as local Jerusalem customs were so apt to do. And as for its peculiar position, before the hymn instead of after the sermon, there is a possible explanation in the fact that when litanies were becoming fashionable in the East as a substitute for the old prayers of the faithful — in the fifth century — the Jerusalem church was precluded from making use of this, the latest liturgical novelty, at the position normal in other rites, by the fact that it had long ago transferred these particular prayers to a point after the consecration. Whether this be the right explanation of affairs at Jerusalem or no, we shall find that when Pope Gelasius at Rome (a.d. 492-6) wanted to get rid of these same antique ‘prayers of the faithful,’ and at the same time wanted to take over the new fashion of litanies, he did adopt precisely this expedient of inserting it after the entrance-chant, just where it stands in the liturgy of S.’James.

3. The Hymn. The equivalent of the group of psalms (/?) before the lections in the Oriental scheme (I) is found in the Greek scheme (III) as another ‘prose hymn’ (3). In the Greek rites this is the Trisagion, the words ‘Holy God, Holy Mighty, Holy Immortal, have mercy upon us,’ repeated three times to a particularly noble melody. This hymn is said to have been divinely revealed (variously to a boy, or a presbyter or the patriarch himself) at Constantinople in the time of the patriarch S. Proclus (a.d. 434-446) as the authentic text of the hymn sung by the angels in heaven. Whatever we may think about this, we have the contemporary testimony of Proclus’ banished predecessor, the heretical ex-patriarch Nestorius, that it was inserted into the liturgy at Constantinople between a.d. 430 and 450. It had been adopted at Antioch before a.d. 471 when the monophysite patriarch Peter ‘the Fuller’ caused a great commotion by adding the clause ...’ immortal, Who wast crucified for us, have mercy...’ and thus turned this Trinitarian hymn into a proclamation of the monophysite doctrine of Christ’s single Divine Nature. In this interpolated form it was adopted by all the Syrian and Coptic Monophysites, who at some time have transferred it from before the lections to a place among the chants between the epistle and gospel. The East Syrians (Nestorians) had adopted it in the Greek position before the eighth century, when it is mentioned by the Nestorian Abraham bar Lipheh. I cannot trace the date of its adoption by the Armenians.

The Greek Introduction is thus made up of elements from two centers, Jerusalem and Constantinople. But it is a scheme as clearly marked as the Oriental and Egyptian schemes, and has spread even more widely than its two rivals. The general trend of influence in liturgical history is always from East to West. The Egyptian Introduction has spread to the Latin churches but not eastwards; one item of the Oriental scheme has spread westwards into the adjacent Greek churches; the Greek Introduction has been copied among the Latins. Only the influence of Jerusalem has been strong enough to work against this current, and spread some marks of the Greek Introduction into the Oriental area.

a. At Rome. The history of the Roman rite is better documented in the fifth-sixth centuries than that of other Western rites, and since Rome exercised an influence of its own in the West, it is convenient to begin with that. When we look at the developed Roman Introduction: (i) Introit or Entrance Chant, (2) Litany, later replaced by the Kyries, (3) Hymn (Gloria), followed by the Greeting and Prayer or Collect, it is clear that it consists structurally of the Greek Introduction (III) followed by the Egyptian one (II) as a sort of double prelude to the lections. It might even be made to appear that the Egyptian (II) scheme was added in the later fourth century and the Greek Introduction (III) prefixed to that during the fifth. This is a neat solution, and may even be true, though it depends on the date of the institution of the collect at Rome. Probst and others have attributed this to Pope S. Damasus (a.d. 366 384) But the documents hardly bear out this tidy idea of the development of the Roman Introduction when the evidence for each item in it is examined separately.

1. The Introit. That erratic document the Liber Pontificalis says of Pope Celestine I (a.d. 422-432) that ‘He ordained that the 150 psalms of David should be sung antiphonally by all before the sacrifice, which used not to be done, but only the epistle of blessed Paul used to be read and the holy gospel.’ The singing of the entire psalter by the congregation at one session before the eucharist can hardly be what is meant; and Duchesne interprets this as referring to the first beginnings of the public recitation of the divine office in the Roman basilicas (as distinct from semi-private services in the oratories of the Roman monasteries). It seems to me that the psalm-chant here described is something much more closely connected with the eucharist than that; certainly it is ‘before the sacrifice,’ but no more so than the epistle and gospel in the compiler’s eyes, i.e. it refers to the institution of a chanting of psalms in the synaxis. The tract and gradual between the lections are certainly older than this, and there remains only the introit, the psalm which by the sixth century was certainly customarily sung at Rome to cover the Pope’s processional entry. The entrance procession had been adopted in the East in some churches at least sixty years before. The chanting of psalms at the eucharist was being extended from the old chants between the lections to other parts of the rite in Africa (though we hear nothing of an introit there) some years before this (Cf. p. 492). The adoption of the procession would appeal to the Roman sense of dignity; and some sort of accompanying chant would hardly be long in making its appearance, just because of the slightly depressing effect of silent processions.

2. The Litany. Alcuin has preserved for us the text of a Latin litany which he styles Deprecatio Gelasii (The Intercession of Gelasius). It is manifestly based on an Eastern model, but Edmund Bishop has shewn that it is undoubtedly of local Roman manufacture in the details of its phrasing, and that there is reason to accept the attribution to Gelasius. Quite recently Dom B. Capelle has pointed out that down to the time of that reputed reformer of the Roman rite Pope Gelasius (a.d. 492-6), the intercessions are frequently referred to at Rome as coming at the end of the synaxis in the old form. After his time they completely disappear at that point except in Holy Week (when they might very well keep their old place as a climax to the synaxis celebrated without the eucharist, to avoid ending abruptly with the sermon). In the sixth century a litany was certainly employed in the Introduction at Rome. These coincidences are too numerous to be accidental. Though the Liber Pontificalis says nothing about it in its vague notice of Gelasius’ liturgical innovations — but then it says nothing of his work upon the Roman eucharistic prayer either — it seems that Gelasius inserted the litany into the Roman Introduction.

It still retained the form of an Eastern litany, with responses said by the people to petitions by the deacon (or by the choir), at least on occasions, down to the time of Pope S. Gregory the Great (c. a.d. 600). But changes were made by him, or more probably had already begun before his time. Writing to the bishop of Syracuse in self-defence against the charge of Western purists that he had followed the customs of Constantinople in the changes he had recently made in the Roman rite, S. Gregory says: ‘We neither used to say nor do we say Kyrie eleison as it is said among the Greeks. For among them all (the congregation) sing it together (as a response to the deacon). But with us (something) is sung by the choir and the people answer it (a populo respondetur). And Christe eleison which is never sung by the Greeks is (at Rome) sung as many times (as Kyrie eleison). But on non-festal days we omit certain things usually sung (i.e. the petitions) and sing only Kyrie eleison and Christe eleison, so that we may spend somewhat longer on these words of supplication.’

Whether the omission of the litany on ordinary days had begun before S. Gregory’s time we cannot say for certain, because the Gelasian Sacramentary, our chief clue to the Roman rite in the century before S. Gregory’s reform, does not contain an Ordinary or outline of all the invariable parts of the rite, but only the text of the prayers said by the celebrant. But it is probably significant that in its rubric directing the omission of the synaxis at the reconciliation of penitents on Maundy Thursday (a ‘non-festal’ mass which would in any case not include the Gloria) the Gelasian Sacramentary simply directs ‘On this day there is no psalm (introit) nor greeting, that is he does not say "The Lord be with you" ‘ (but begins straight away with the prayers for the penitents). But at the baptismal eucharist at the end of the paschal vigil (the festal mass of Easter) we find ‘Then while the litany is sung (the bishop) goes to his throne, and intones Glory be to God on high.’ Here the litany seems plainly to be as much a feature of the rite reserved for festivals as is the Gloria. And there are no Kyries between Introit and Greeting-Collect at non-festal masses like that of the penitents on Maundy Thursday because S. Gregory had not yet invented them.

What S. Gregory’s work on this part of the Roman rite seems to amount to is this: he left the litany on festal days perhaps more or less as it had been before (though it is as well to note that we have no evidence either way whether or not the text of the Deprecatio Gelasii as preserved by Alcuin still represented the current usage at Rome at the end of the sixth century). On non-festal occasions S. Gregory instituted repetitions of Kyrie eleison and Christe eleison between the introit and the greeting and collect, where previously nothing had intervened on non-festal days when the litany was not said. Though there is no direct evidence on the point I see no reason to doubt that these as S. Gregory fixed them numbered nine (with Christe eleison for the middle three) sung alternately by the choir and people.

But S. Gregory’s innovation of the Kyries used as a chant instead of the litany on non-festal days soon ousted the use of the litany on festivals also. The text of the litany as used on festivals has left no trace in any extant MS. of the Gregorian Sacramentary. The litany, greatly developed and in some things transformed, continued to hold a place in Roman usage, but as an almost separate rite conducted in procession through the streets outside the church as a preliminary to the eucharist on days of solemn supplication. It was thereby enabled to survive as an actual part of the Roman eucharistic rite at the Easter and Whitsun baptismal masses and (transferred to a later point in the rite) at ordinations and monastic professions. In this form, as a solemn supplication, it was adopted in France for occasions like the processions of the Rogation Days, at first as an addition to and then instead of the old French ‘procession’ of penitential psalms, certainly before the end of the eighth and perhaps before the end of the seventh century.

3. The Hymn. We have seen that before the time of Pope Celestine I nothing whatever preceded the lections; and even after he had introduced the introit it formed the whole of the Roman introduction, according to the Liber Pontificate. This was probably the Constantinople and Jerusalem usage of the time, and lasted at Rome from c. a.d. 430 till the introduction of the collect, of which we shall say more in a moment. Jerusalem may have introduced the litany after the entrance chant quite early in the fifth century; Rome certainly followed suit at the very end of the century. Meanwhile Constantinople had introduced the hymn of the Trisagion between the entrance chant and the lections (before a.d. 450), and Antioch (and probably Jerusalem) had done the same before c. a.d. 470. Once more Rome followed the Eastern custom, but after a generation or two. Pope Symmachus (a.d. 498-514) ‘Ordained that on every Sunday and martyr’s feast the hymn "Glory be to God on high" should be said.’ Both the position of this hymn and the frequent Roman description of it as ‘the Angels’ hymn’ witness to its relation to ‘the Angels’ hymn’ of the Trisagion at Constantinople. The Eastern structure of S. James — (1) Eiso-dikon (= entrance chant, Monogenes), (2) litany, (3) Trisagion (‘hymn of the Angels’) is exactly reproduced by the Roman (1) Introitus ( = entrance chant,’ a psalm), (2) litany, (3) Gloria (‘hymn of the Angels’). The Roman church refused to change its old scriptural entrance-chant of a psalm for the new Greek Monogenes, composed by an emperor of dubious orthodoxy; and likewise substituted Gloria in excelsis as the scriptural ‘hymn of the Angels,’ to avoid being committed to the apocryphal legend of the divine revelation of the Trisagion. But it adopted the whole (III) Greek structure of the Introduction — entrance chant, litany, hymn — nevertheless, though it did so only item by item.

The Gloria was no new composition when it was put to this new use at Rome c. a.d. 500. It is found in Egypt, Syria and Asia Minor in the fourth-fifth century, and is said to have been introduced into the West by S. Hilary of Poitiers c. A.D. 363, who had come upon it during his banishment in the East. The number of local variants in the text of the hymn already found in the fourth century indicate an origin in the third, or even perhaps the second century. It had been a pre-Nicene Eastern ‘hymn at dawn,’ and thus found its way into the new public morning office of Lauds in the East, where it formed a sort of ‘greater doxology’ at the end of the psalmody. In this position the Roman church seems always to have employed the Benedictus or Song of Zachariah. The Gloria, the old hymn which began with the song of the angels at Bethlehem, was therefore available at Rome for use at the eucharist, when current fashion suggested the need of an ‘Angels’ hymn’ before the lections of the synaxis. There could indeed be no more suitable text than this to celebrate the redemption which the scriptures announce.

But it is perhaps a symptom of the reluctance with which the Roman church accepted innovations which had not an obvious practical purpose (like the introit), that both the litany and the hymn, which in the East became at once fixed and unvarying parts of the rite whenever it was celebrated, were adopted at Rome only as ‘decorations’ suitable to elaborate it for festivals, but not integral to the real purpose of the liturgy. This ‘occasional’ use may, too, reflect the growing influence of the calendar on the Western rites, which gives rise to the use of variable prayers in the West during the fifth century, an innovation which the East did not adopt in that form. But there is also something of the Roman concentration on the main purpose and end of the liturgy and the sense of its form (which comes out again in the directness and brevity of the Roman prayers) about this reluctance to amplify the rite on all occasions with purely decorative additions. It seems indeed to have been felt at Rome that a hymn at this point was suitable even on feast-days only at the specially solemn ‘stational’ eucharist of the bishop. It is mentioned only once in the Gelasian Sacramentary at the Easter vigil (when any celebrant might use it). But the Gregorian Sacramentary, though it follows Pope Symmachus’ ruling that the Gloria was to be used on Sundays and feasts, restricts this to the stational eucharist celebrated by a bishop for his whole church. Presbyters are permitted to use it only on Easter Day (to which later custom added the anniversary of a priest’s own ordination). It was only in the eleventh or twelfth century that priests began to use it on all Sundays and festivals like bishops. The omission of the Gloria on Sundays in Advent and from Septuagesima to Easter is not indicated in the Gregorian Sacramentary, but is suggested by the Ordo Romanus Primus, where it is used si tempus fuerit ‘if it is the season for it.’ This further restriction in the use of the hymn (which is not found in the Eastern use of the Trisagion) may not have suggested itself until the seventh century.

4. The Greeting and Prayer. The synaxis on Good Friday in the Roman missal — the only really ancient specimen surviving of the old form of the Roman synaxis — opens abruptly, without introit (or of course Kyries and Gloria) and also without a collect. This seems to bear out the statement of the Liber Pontificalis that when Celestine first prefixed the introit no other text intervened between it and the lections. It is true that modern liturgical scholars have almost unanimously attributed the origin of the Roman collect to S. Damasus fifty years before Celestine. But this question is so closely bound up with the whole problem of the origin of prayers varying with the calendar, in the Western rites as a group and not the Roman ritealone, that it seems better to leave it for discussion in this larger setting in the next chapter, and to rest the case for the moment upon the evidence of the Liber Pontificate that there was still no collect in Celestine’s time c. a.d. 430. But if its insertion is later than this, there is reason to think it appeared not very much later, say within the next twenty or thirty years.

From the fact that the greeting at Rome is placed before this prayer, as in the Egyptian rite (and not immediately after the entrance chant as at Constantinople, or in its original place immediately before the lections as in Spain), we may be justified in supposing that the custom of a prayer before the lections was borrowed at Rome from Alexandria; and we do in fact find that from c. a.d. 430-445 relations between the Roman and Alexandrian churches were closer than at any other time between the visit of Athanasius to Rome in a.d. 339 and the last rapprochement of these two sees in the time of S. Gregory the Great c. a.d. 595. But it must be noted that while the Alexandrian collect of the fourth century is an unchanging prayer, the same on all occasions, the Roman collect when we first meet it is already one which varies with the occasion. There may have been a period when the Roman collect also was unvarying and referred simply to the hearing of the scriptures, like the Egyptian ones. But if so, this period must have been short, for it has left no trace whatever in the Roman evidence.

The following seems, then, to be the approximate history of the Roman Introduction to the synaxis. Celestine I prefixed the introit, the chanting of a psalm during the entrance procession, c. a.d. 430. Before that time there had been no Introduction whatever at Rome before the lections. In the next twenty years or so the Egyptian Introduction (II) of a Greeting and Prayer was set between the introit and the lections. There must after that have been a period when the Roman Introduction consisted simply of introit, greeting and collect, followed by lections. This is precisely the arrangement still implied for non-feast days in the first rubric of the sixth century Gelasian Sacramentary cited on p. 453. At the very end of the fifth century Gelasius added the litany, probably from the rite of Jerusalem, between the introit and the greeting. A few years later Symmachus again added the hymn between the new litany and the greeting. Perhaps the litany, and certainly the hymn, were from the first special to Sundays and feasts. They were placed where they were to avoid disturbing the Egyptian ‘group’ of greeting and prayer; which suggests that the Egyptian idea of the prayer as specially connected with the lections immediately after it had at one time obtained a foothold at Rome. The whole Roman Introduction is therefore a product of the period between c. a.d. 430 and c. a.d. 500, precisely the period when we have seen that the adult catechumenate was ceasing to be of any practical importance at Rome. The Introduction at Rome represents, therefore, the adaptation of the old pre-Nicene synaxis, which had had to serve the purposes of propaganda outside the ranks of the faithful, to the needs of a ‘public’ worship in the new Christian world.

b. The Western Introduction outside Rome. We are on much less secure ground in dealing with the Western rites other than that of Rome right down to the seventh, and in many matters the ninth, century. Before then the evidence available is both less in quantity and more ambiguous in quality than in the case of the Roman rite; and the subject is encumbered with modern theories, no one of which seems to account for all the facts. We shall not enter upon them, but merely note what evidence is available, and what it indicates.

It is necessary at the outset, however, for the sake of those who have read the usual manuals, to take account of two modern discoveries which seriously alter the bearing of this evidence. Dom Wilmart’s demonstration — the word is not too strong — that the so-called Epistles of Germanus of Paris (d. a.d. 576) have nothing to do with Germanus or Paris, but were composed in the South of France (or perhaps in Spain) c. a.d. 700, will necessitate a considerable reconstruction of what one might call the ‘usual’ theory — though in fact it is mainly of French construction — of the history of the ‘Gallican’ rite. The term ‘Gallican’ was first used to cover only the old local rites used in some parts of what is now geographically France before the end of the eighth century. These rites have someclear resemblances to the Spanish Mozarabic rite. Successive French authors — Martene, Le Brun and above all Duchesne — grounding themselves on these resemblances, and noting parallels real or supposed in other Western rites, and assuming always that the ‘French’ rites were the parent, or at all events the purest representative, of the whole group, have extended the term ‘Gallican’ to mean in practice ‘all Western European rites other than that of the city of Rome.’ Not content with thus stretching the meaning of the term, some disciples of this school speak of ‘the Gallican rite’ as originally observed throughout the whole West including Africa, leaving the Roman rite as an isolated enigma confined to the city and suburbs of Rome. Upon analysis, it will be found that the key-point of the theory is always the Letters of Germanus. Now that these turn out to be at least as much Spanish as French, and to represent ‘Gallicanism’ not in its early purity but in the period of its admitted decadence after it had been transformed by a number of foreign elements, the term needs to be handled with more caution. We arc thrown back on the older genuinely French evidence for the French rites, which is less abundant than one could wish. To avoid begging any questions the word ‘Gallican’ ought to be used only in its original sense of rites which existed within the geographical boundaries of what is now called France, which was then neither a racial nor a political nor an ecclesiastical unity. (When it is used in the wider modern sense of ‘Western but not Roman’ it will henceforward be placed in inverted commas.)

The second fact of which account must be taken is Dom Connolly’s vindication of the authorship of the treatise de Sacramentis for S. Ambrose of Milan c. a.d. 400. Here at the end of the fourth century Milan is already using what is recognisably an early form of the Roman canon. It means that the Milanese rite is fundamentally a Roman — or as I should prefer to put it, an Italian — rite, which in the course of later history has received some ‘Gallican’ decorations, and not an originally ‘Gallican’ rite which has been subsequently Romanised. With the recognition of this we must abandon forthwith Duchesne’s theory that Milan was the centre of diffusion for the ‘Gallican’ rite in the West, whither he supposed it had been imported from the East by Ambrose’ oriental predecessor, Auxentius, c. A.D. 360. With the elimination of the theory of an oriental origin for all non-Roman Western rites the greatest single unnecessary obstacle to a clear understanding of the development of the eucharistic liturgy in the West is removed.

So much by way of general preface to the special question of the Introduction to the synaxis in the West outside Rome. We shall return to the larger issues later; here the facts are these:

At Milan. We know virtually nothing of the development of the Milanese rite between the late fourth century (in de Sacramentis) and the ninth, when its text comes into view in the Sacramentary of Biasca. The Introduction is arranged thus: 1. The Ingressa, a psalm chant analogous to but not identical with the Roman introit. 2 (a). A diaconal litany, which like the Roman Deprecatio Gelasii is based on an Eastern text, but not identical with the Roman version. There are two forms of this litany at Milan, one used on the first, third and fifth Sundays in Lent, and the other on the second and fourth. It is not used at other times. Or 2 (b). When the litany is not used at Milan there is a hymn, consisting of Gloria in excelsis (in somewhat expanded form). 3. After the litany or the Gloria, there always follows Kyrie eleison repeated thrice. 4. After the Kyrie follows the greeting and the collect as at Rome (and then the lections).

This differs from the Roman Introduction, a. in making the litany and hymn alternatives; b. in the insertion of the threefold Kyrie after the hymn, or after the litany in Lent.

a. The atrophy of the litany seems to have taken rather different forms at Milan and Rome. At Rome it disappeared altogether, replaced by the ninefold Kyrie, first inserted as an alternative to it by S. Gregory c. a.d. 595. At Milan it survived in Lent, as a special observance.

b. The Milanese threefold Kyrie does not seem to be any sort of survival of a litany, despite all that has been said to that effect by French scholars. There are no petitions by the deacon, and no trace that there ever were any. On the contrary, the threefold Kyrie is appended to the Milanese litany when it is said, just as it is to the Gloria at other seasons. Musically, it is treated as a hymn. A similar threefold Kyrie as a hymn is found in some of the French rites, where it goes back to the Council of Vaison in 529, which instituted it in France, in imitation of ‘a custom which has been introduced both in the Apostolic See and in all the Eastern and Italian provinces.’ ‘Italy’ means at this time what we call ‘North Italy’ — the region of Milan. The Milanese Kyrie is therefore not a ‘Gallican’ feature imported into the Milanese rite, but something which existed at Milan before the French rites borrowed it. It seems in fact to be the original Milanese form of ‘the hymn’ before the collect. We do not know when it first came into use there for this purpose, but it seems (from the phrase intromissa est used at Vaison) to have been supposed to be fairly new everywhere in a.d. 529, i.e. the Kyries were adopted at Milan about the same time as the Gloria at Rome. Both met a need instinctively felt for a ‘hymn’ before the lections, to adapt the old synaxis form to the new conditions. When, later on, the Roman hymn spread northwards, the native Milanese equivalent, the threefold Kyrie, was short enough to be added to it instead of being displaced.

So far as its Introduction goes, therefore, the Milanese rite developed under much the same conditions as the Roman rite, and in the same period. It shews later signs of the influence of the Roman rite to the south of it during the sixth-ninth centuries, just as naturally as it shews other signs of the influence of its other neighbours, the Gallican rites to the north-west of it, during the same period. Eucharistic rites never have existed in water-tight compartments or rigidly excluded each other’s influence. On the contrary they have borrowed freely from one another in all ages down to the sixteenth century, and this even across the barriers erected by open breaches of ecclesiastical communion. The Milanese rite in its basis is neither French nor Oriental but Italian, like the Roman. And like the Roman rite it has had its own local history within the general Italian setting, which has left its marks upon its modern form. All things considered, this account of the matter is only what might have been expected.

In Spain. The exact history of the Spanish Introduction is not very easy to make out, but the following are the main facts. The Mozarabic Introduction is as follows: 1. The Antiphona ad praelegendum, (usually) a psalm-chant, corresponding to the introit. (2. On great feasts, an interpolated version of the Trisagion, the interpolations varying according to the day.) 3. On Sundays and all feasts Gloria in Excelsis. 4. The collect. 5. The greeting. 6. The lections. We are handicapped as to the history of the different items by the fact that neither of the two earliest known Spanish MSS. — the Antiphoner of Leon (ninth-tenth century) which contains the chants and the music of the rite, and the Sacramentary of Toledo (ninth century) which contains the prayers — is equipped with an ‘Ordinary’ of the rite as a whole. Furthermore, it is uncertain just how old the arrangements are to which either of these MSS. witnesses. Both of them are substantially copies of older MSS. going back to the later seventh or eighth centuries. But it is possible (a) that one or both of the extant MSS. have to some extent been brought up to date, to conform to current custom when they were written, (b) that in some things this was not done, and that they witness to a state of affairs which was obsolete or obsolescent in the ninth century. The following facts are to be noted:

1. On all fast days the modern Mozarabic mass begins without any introduction at all, but simply with the greeting and lections, like the primitive rites. In the Antiphoner of Leon, however, the Antiphona is always said, even on fast days, unless the office of None has just been said in choir (when there would naturally be no entrance-procession, since the clergy would be already in church). This appears to witness to two stages: a. A period when there was no Introduction at all beyond the preliminary greeting, as at Rome before Celestine, c. a.d. 430 (Cf. p. 453). (It is noteworthy that the African rite, which has been supposed to have some affinities with that of Spain, seems never to have developed an Introduction at all.) b. A period when the Introduction consisted only of entrance chant (Antiphona) followed at once by greeting and lections, as at Rome in the period immediately after Celestine’s innovation (Cf. p. 458).

2. The variable Trisagion on great feasts is evidently an instance of that growth of Byzantine influence which followed Justinian’s reconquest of part of Spain in the sixth century. How soon it was interpolated into the rite after that date it is impossible to say. The first evidence of its use (on four days in the year) is in the nin± century Antiphoner, to which some eleventh century MSS. add three other days. In the earliest MS. which gives any sort of ‘Ordinary’ (Toleten. 35, 4) of the tenth century, it is ignored altogether, but this is not unnatural in the case of an exceptional festal feature of the rite. All things considered, it may well have formed part of the late seventh century arrangements which were copied into the ninth century Antiphoner, but it is hardly likely to be much older than that. It is noticeable that S. Isidore in his description of the Toledan rite in the early years of the seventh century does not mention it.

3. The first reference to the use of Gloria in excelsis in the Spanish mass ‘on Sundays and all feasts’ appears to be in the late eighth-century writing of Beatus of Liebana and Etherius of Osma adv. Elipandum. It is found (apparently with three different musical settings) in the ninth century Antiphoner,’ in one of which the wording contains variants somewhat akin to those of the Milanese version. Its use is evidently borrowed from Italy as the use of the Trisagion is borrowed from the East, perhaps at about the same time, though it became the normal hymn of the Introduction, while the Trisagion was an occasional extra for special feasts. The Roman rite as a whole was in use in some parts of Spain, e.g. Galicia, in the later sixth century, which rnay have led to the adoption of the Roman hymn in other places which were properly Mozarabic in rite. But such early seventh century references to the Gloria as I have found in Spain all seem to refer to its use at Lauds (e.g. can. 12 of the fourth Council of Toledo a.d. 633).

4. The Collect. There is some contradiction in the evidence about this. S. Isidore in the early seventh century says nothing of a collect before the lections, but specifically calls the Missa after the sermon the ‘first prayer’ of the rite. The ninth century Lib. Moz. Sac. likewise makes no provision for what we should call a ‘collect’ at all, and though some of the masses in the eleventh century Liber Ordinum have a variable collect, others, like those of the ninth century Sacramentary, are still without any prayer in this position. It would appear from all this that the variable collect made its first appearance in the Spanish rite surprisingly late — in the tenth-eleventh century.

The only difficulty in accepting this account of the Spanish Introduction is a little rubric in the Antiphoner, which orders that on Palm Sunday after the Aniiphona ad praelegendum ‘Kyrie eleison is not said’ (as though it were said on other occasions) ‘but the bishop forthwith says the collect, and after the collect there follow the lections.’ This looks as though the Roman Introduction (of penitential seasons), introit, Kyries, collect, lections, were the normal thing in some Spanish churches when the Antiphoner was compiled. Whether this was the case or not, I am unable to say; but I know of no other evidence for it, or for the use of the Kyries at all in the Spanish Introduction.

If we may ignore this tantalizing statement, the history of the Spanish Introduction appears to be approximately as follows: It begins, like the Roman, as an entrance-chant followed by the greeting and lections. Perhaps in the late sixth century, more probably in the seventh, the Roman ‘hymn’ was inserted on Sundays and festivals, supplemented on great feasts by the Byzantine one. This formed the whole Introduction down to the tenth century. Then the use of the variable collect before the lections was taken over from the other Western rites; but it was attached in thought to the Introduction which preceded it rather than to the lections which followed it, as in the Roman idea. The greeting therefore was left preceding the lections in the primitive position, to mark the break between them and the Introduction, and not placed before the prayer as in the Egyptian and Roman rites.

The Gallican Rites. We are now in a better position to approach the real difficulty in discerning the development of the Western Introductions — the French evidence. Deprived of the delusive certainties of ‘Germanus,’ our information has to be pieced together from various sources, always a process which offers plentiful opportunities of error.

‘Germanus’ presents us with the following elaborate opening: 1. The Antiphona, an entrance-chant (a psalm?). 2. Greeting. 3. Trisagion (which he caUs by its Greek name, Aius). 4. A Kyrie-hymn, like that at Milan. 5. The Benedictus (Song of Zachariah). 6. The O.T. lection and Epistle. 7. Benedictus es (Song of the Three Children). How much of all this can we verify from other sources?

1. No other early Gallican document offers any evidence of such an entrance-chant, and since ‘Germanus’ calls it by its Spanish name Antiphona, we may perhaps dismiss it from the original Gallican rite as a later Spanish importation. 2. The Greeting. What is noticeable is that this is placed in the Antiochene position immediately after the entrance, and not as in Italy before the collect; or as in Spain, before the lections. ‘Germanus’ has no reference to a collect in the Introduction at all, though the Gallican evidence of the seventh-eighth century places one after the Benedictus. Taken in conjunction with the absence of a collect from the Spanish rite down to the ninth-tenth century, this omission in ‘Germanus’ is significant of its ‘Hispanising’ tendency.

One notes next the collocation of the opening Antiphona with a version of the Trisagion, as in Spain. The group of three successive chants Trisagion, Kyrie and Benedictus, seems elaborate, but here ‘Germanus’ begins to make contact with other Gallican evidence. The Bobbio Missal (seventh-eighth century) which, though probably compiled originally by an Irishman and written in Italy, contains a great deal of French material, makes provision in the Introduction for the Trisagion (which it calls Aios), the Gloria in excelsis, the Benedictus (which it calls Prophetia) and a deacon’s Litany (which it calls Prex or Preces). It places them in that order, but does not specify the way in which they are to be fitted into the rite. The Gloria and the Litany come, however, from a Celtic Ordinary, as we shall see; so that we are left with Trisagion and Benedictus in that order, as the compiler’s idea of the Gallican Introduction, as in ‘Germanus.’ (Probably in arranging for all four chants the compiler of Bobbio is trying to make his book do for churches which used either system, though he does not say that they are alternatives.)

We have, however, an earlier reference to the Trisagion in Gaul, in the almost contemporary life of S. Gaugericus, bishop of Cambrai c. A.D. 600. It is to be noted that while the Spanish books have their Trisagion in the orthodox form, the Bobbio Missal plainly implies that it expects it to be sung in the Gallican use with the Syrian monophysite interpolation ‘Who wast crucified for us.’ Furthermore, there is a threefold Kyrie eleison after the Trisagion in the Syriac S.James, as in ‘Germanus.’ Taking this in conjunction with the Antiochene greeting immediately before the Trisagion in ‘Germanus,’ it seems fairly easy to see whence the model for all this part of the ‘Germanus’ rite in its present form was derived — from Syria.

The Kyrie-hymn is appended to the Trisagion in Gaul as it is appended to the Gloria at Milan, and probably for the same reason — that it is the original opening chant, dating from can. 3 of the Council of Vaison in a.d. 529. The Trisagion was imported into Gaul from Syria later in the century, but the native hymn was brief enough to survive as an appendage to the new importation. We do not hear of the Trisagion in Gaul until the very end of the sixth century, which is the period when evidences of the importance of Syrians in Gaul are most numerous.

The evidence for the use of the Benedictus in the Gallican rite is solid and satisfactory. Two collects in the Burgundian Missale Gothicum (eighth century), seven in the Bobbio Missal (seventh-eighth century) and two in the oldest extant Gallican missal, the Masses of Mone (seventh century) are all plainly intended for use after the Benedictus. Gregory of Tours in the sixth century speaks of the bishop intoning the Benedictus at an early point in the liturgy at Tours. It evidently held the place in the sixth century French rites that Gloria in excelsis held in the sixth century Italian rites, as the ‘hymn’ before the collect. Its use in place of the Gloria is probably due to the fact that in the sixth century the Gloria in France was used at Lauds in the place where the Italian office-books used the Benedictus.

Little can be inferred from the use of the ‘Song of the Three Children’ after the epistle in ‘Germanus.’ The true Spanish place for it was before the epistle, between that and the preceding O.T. lesson, but this was not always adhered to. The Gallican lectionary of Luxeuil mentions it twice — once after the O.T. lesson and once after the epistle — which does not help. It is found in the Roman rite after the last O.T. lesson on Ember Saturdays, so that its use is common to all the Western rites. Gregory of Tours mentions it only at Mattins in Gaul.

We find in Gaul, therefore, c. a.d. 600 an Introduction consisting of (1) the Trisagion, (2) the threefold Kyrie, (3) Benedictus, (4) greeting and collect. It is obvious that the developed Gallican structure is precisely the same as the developed Roman one — (1) entrance-chant, (2) Kyries, (3) hymn, (4) greeting and collect, though the texts used are not the same and the Kyries in Gaul are older by three-quarters of a century than at Rome. It is further noticeable that the .Gallican rites of the seventh (and presumably the sixth) century, have the greeting and a variable collect immediately before the lections — a feature which did not yet exist in the Spanish or African rites, or indeed in any but the Italian ones. The coincidence can hardly be accidental. The Roman Introduction, completed by the end of the fifth century, was known and deliberately imitated by the French churches of the sixth century, even though the imitation was by no means servile.

The question now seems legitimate — was S. Gregory in instituting the ninefold Kyrie at Rome influenced rather by the use of the threefold Kyrie-hymn at Milan and in Gaul than by any reminiscences of the Roman litany already obsolescent as a normal feature of the Roman rite? For what its evidence is worth for Roman practice in the sixth century, the Deprecatio Gelasii witnesses that Kyrie eleison was not the old Roman litany-response, but the Latin phrase Domine exaudi et miserere.

The Celtic Introduction. The Bobbio Missal, as we have said, has a ‘mixed’ Introduction which sets Irish and Gallican elements side by side. The subtraction of the latter leaves an ‘Ordinary’ or outline of the rite almost identical with that found in the pure Irish Stowe Missal. This latter was copied c. a.d. 800 from an older Irish MS. written not later than c. a.d. 650 and probably somewhat earlier. The common elements of the Bobbio and Stowe books present us, therefore, with the Irish rite of the first half of the seventh century.

This earliest known Irish rite is recognizably the Roman rite both in structure and contents. It is, of course, ‘Roman’ in the usual Irish way, both old-fashioned and curiously embellished, for Ireland was a long way off and Irish scribes were inveterate and often wayward ‘improvers’ of the texts they copied, whose taste in things liturgical was always for the unusual. But apart from such ‘tinkerings’ (as Edmund Bishop was wont to call the Irish way with liturgical documents) the Irish rite is Roman not only in substance but in eighty per cent of its details.

The Introduction in Stowe is as follows: 1. A collect (drawn from the Gregorian Sacramentary) is sung: ‘O God Who having confided unto blessed Peter Thine apostle the keys of the kingdom of heaven didst bestow on him the pontifical office of loosing and binding souls: mercifully receive our prayers; and by his intercession we entreat Thee, O Lord, for help that we may be loosed from the bondage of our sins; through...’ There follows 2. Gloria in excelsis, 3. the greeting and collect, 4. the epistle, followed by 5. a gradual chant and a deacon’s litany, which is related to the Deprecatio Gelasii (i.e. it seems to be an independent translation and re-working of the same Greek original). The Celtic Introduction when we first meet it thus consists of 1. a Roman prayer, 2. the Roman hymn, 3. the Roman variable collect — with a litany similar to the Roman one, but after the epistle instead of before the hymn. There is not much doubt of where the materials of the Irish rite were drawn from, even if Stowe did not professedly give them as the ‘collects and prayers of the mass of the Roman church.’ Though evident traces of S. Gregory’s reforms of the Roman rite are to be found in both Stowe and Bobbio, both books preserve details of the pre-Gregorian Roman rite, notably in some readings in the canon. It is conceivable that what we have is a revision of an older Irish version of the Roman rite, brought into line c. a.d. 620-650 with the recent Gregorian reforms.

Conclusion. This has had to be a lengthy and somewhat technical consideration, but it has enabled us to clear up a series of problems which have evidently given rise to much perplexity in the minds of all the compilers of liturgical histories and text-books. The facts appear to be as follows: The original nucleus of the synaxis sufficed the church as long as she existed in a heathen world and for a generation or two afterwards. When the world at large began to turn towards Christianity and the synaxis began to need adaptation to a public worship, three different schemes of Introduction arose in the East, which had all found their final form before the end of the fifth century. The same needs were felt in the West, but development there was rather slower. The Roman Introduction which combined the Greek and Egyptian schemes was built up piece by piece between c. a.d. 425 and 500, and the Roman scheme thus formed was the basis of the other Western schemes. The frequency with which we find that it was borrowed without the Gelasian litany suggests that it spread chiefly in the later sixth century, when it appears that the litany was dropping out of regular use at Rome itself. It can be suggested that, in return for the Roman outline of the Introduction, the other Western churches contributed to the Roman rite the Kyrie-hymn with which S. Gregory replaced the Roman litany; though S. Gregory gave it a local Roman adaptation in making it ninefold instead of threefold, and inserting Christe eleison. Just so the other Western rites adapted the Roman Introduction to some extent when they took it over. So in the same way we find that in adopting elements of the Eastern Introductions the Roman and other Western churches freely exercised their own taste and judgement.

All over Christendom the addition of the Introduction was intended to serve the same purpose — to strengthen the element of worship in the synaxis, once the decline of the catechumenate had removed the restriction on this caused by the presence of non-christians. It is thus natural that the only item of the Introduction which is found in all rites in some form is the ‘hymn’ before the lections, whether it be drawn from the Psalter as in the ‘Far Eastern’ rites, or is in the form of a ‘prose hymn’ as in the Greek and Western rites.

Though the order in which lections from the various parts of the Bible were read was already fixed in pre-Nicene times (Cf. p. 39), there appears to have been no such general agreement then as to the number of lections which should normally be read at the synaxis. The absence of other elements than lections (with the accompanying chants and sermon) gave time for a relatively large number of passages to be read without unduly prolonging the service. This multiplicity of the pre-Nicene lections continued in some churches in post-Nicene times, especially in Syria. But towards the end of the fourth century the growth of other elements in the synaxis brought about the limitation of the lections in most churches to three, (1) from the O.T., (2) the apostolic writings and (3) the gospel, as a normal rule. In Africa, Spain and Gaul, and perhaps in some other churches, it was then customary on martyrs’ feasts to substitute for the O.T. lection an account of the martyr commemorated on that day; and in some churches lections from apocryphal ‘apostolic’ writings were even substituted on occasions for the second lection from the canonical epistles. The use of uncanonical gospels for the liturgical lessons is attested in the second century, and that of ‘harmonies’ or conflations of the four gospels like Tatian’s Diates-saron (second century) lasted on, especially in Holy Week, to as late as the seventh or eighth century in many churches from E. Syria to Spain. At Rome, however, a rigidly scriptural tradition always prevailed in the matter of the lections, which excluded not only apocryphal writings and ‘harmonies’ but also the historical ‘acts’ of the martyrs from the eucharistic liturgy; though the latter were accepted into the lessons of the Roman office, apparently in the seventh century. One main result of the general spread of Roman influence through the Western churches was the elimination of all non-scriptural lections at the eucharist in the West.

In the fifth century the church of Constantinople began to reduce the normal three lections to two by the abolition of the first (from the O.T.). Rome followed suit in the late fifth or early sixth century, though the process was slower at Rome; the full three lections are still found provided for a few days in the year in the seventh century Roman lectionary list known as the ‘Wurzburg Capitulary.’ Indeed it may be said that the process of ‘dropping’ the O.T. lesson was never completed at all in the Roman rite, since the Wednesday and Saturday Ember Days still retain two and five O.T. lections each in the Roman missal; and on the weekdays of Lent and certain other days it is not the O.T. lesson but the epistle which has vanished. It does in fact not infrequently happen that the aptest comment on a passage of the gospels is furnished not by the New Testament but by the Old. In retaining the liberty of using passages from any part of the Bible in combination with the gospel the sixth century Roman church shewed good judgement, though the subsequent dislocation of the Roman lectionary (Cf.p. 364, n. 1) prevents this wisdom from being always apparent in the modern missal. The omission of the third lection from other rites than the Byzantine and Roman was both later and less usual, though it had begun in many churches by the seventh-eighth Centuries, at least on ordinary days. It is sometimes suggested that the possession of three lections is a characteristic of the ‘Gallican’ rite while two is ‘‘Roman.’ But all rites, or at all events all Western rites, were three lection rites in the early fifth century. The retention of three lections therefore gives no real clue to the origin of a particular rite; it is at the best one indication! of its later history.

The chants which came between the lections have their own history, which is still obscure in certain points, but which need not detain us here. The psalm-chant with Alleluias (gradual), which came down from the synagogues of our Lord’s time was always reserved for the place of honour immediately before the gospel. The invention of Lent in the fourth century led to the suppression of the Alleluias during this penitential season (and of the verse which had been added after them in die Roman rite, apparently from Byzantium, during the seventh century). In their place was substituted the Tract, a psalm-chant which had formerly intervened between the Old Testament lesson and the epistle, the retention of the O.T. lesson during this season apparently leading to the retention of the chant which was regarded as a comment upon it. The Gallican rite made various innovations in the way of elaborating and adding to the chants between the epistle and gospel, of which the latest were the mediaeval Sequences, metrical compositions (not always of a very edifying character) of which five of the best are still to be found in the modern Western rite. But all these changes are characteristically French mediaeval elaborations upon the simple psalm chants, with Alleluias added before the gospel, which had always been interposed between the lections of the synaxis since the time of the apostles. These are still found in every rite of catholic Christendom with one exception. Archbishop Cranmer directly forbade the use of any chant whatever between the epistle and gospel in 1549.

The business of the preacher of the sermon which followed was to expound and interpret the salvation declared in the scriptures which had just been read, as is clear e.g. from the Egyptian prayers before the lections already quoted (Cf. p. 447). The same note is echoed in the prayer after the sermon in Sarapion’s collection, a feature of the Egyptian rites which appears to be unique as a developed formal constituent of the rite.

In the synaxis rite of Sarapion it runs thus:

‘After the rising up from the sermon — a prayer:

‘O God the Saviour, God of the universe, Lord and Fashioner of all things, Begetter of the Only-begotten, Who hast begotten the living and true Expression (of Thyself, charactera, cf. Heb. i. 3), Who didst send Him for the rescue of the human race, Who through Him didst call mankind and make them Thine own possession; we pray Thee on behalf of this people. Send forth "holy spirit" and let the Lord Jesus visit them; let Him speak in their understandings and dispose their hearts to faith; let Him Himself draw their souls to Thee, O God of mercies. Possess Thyself of a people in this city also, possess Thyself of a true flock: Through ...’

Apart from the renewed insistence on the theme of the ‘rescue’ of humanity in Jesus, we may note here the survival of the notion — becoming a little old-fashioned in Sarapion’s day — of impersonal ‘holy spirit’ (without the definite article) as the medium whereby the Lord Jesus ‘visits’ His members on earth and Himself speaks in their understandings and disposes them to believe. Theology in the fourth century was beginning to attribute such operations to the Personal action of the Holy Spirit, but a brief comparison of Sarapion’s expressions with e.g. another Egyptian work, S. Athanasius’ de Incarnatione, will shew that he was by no means alone in still retaining the older attribution to the Logos, the Second Person. His ‘invocation’ of the Logos to supervene in the consecration of the eucharist is quite of a piece with the rest of his theology.

The prayer after the sermon has disappeared from the text of the Alexandrian liturgy of S. Mark (no doubt through the infrequency of preaching in Byzantine times). But it is referred to several times by Origen in his homilies at Alexandria during the third century, and once by S. Athanasius in the fourth. Evidently the rule against praying in the presence of catechumens was differently interpreted in Egypt from the way in which it was understood elsewhere.

In the later fourth century in Africa, and perhaps elsewhere, the place of this prayer was to some extent supplied by a long fixed ‘ascription’ at the end of the sermon. Three of Augustine’s sermons have preserved the full text of this as their concluding paragraph, and the cue for it ends quite a number of others: ‘Turning unto the Lord God the Father Almighty with a pure heart let us render unto Him, so far as our littleness may, most hearty and abundant thanks: beseeching His singular goodness with our whole intent that of His gracious favour He would vouchsafe to hear our prayers; and by His might drive far the enemy from all our doings and thoughts; increase in us our faith, govern our minds, grant unto us spiritual desires and bring us to His everlasting bliss; through Jesus Christ His Son our Lord, Who with Him liveth and reigneth in the unity of the Holy Ghost, God for ever and ever. Amen.’

The way in which this prayer takes the ostensible form of an address or exhortation to the people (known as a praefatio by contrast with an oratio addressed directly to God) is a characteristic of Western rites which we shall meet again. But the sermons of S. Fulgentius of Ruspe, an African bishop a century later than Augustine, end not with an invariable ascription but with a variety of formulae, frequently containing a reference to the feast or saint of the day. It is a little indication of the way in which during the fifth century the ecclesiastical calendar came to exercise an influence over the old fixed prayers of the liturgy in the West, a tendency which had hardly begun in Augustine’s day. S. Leo’s sermons at Rome c. a.d. 450 end with the simple ascription ‘through Jesus Christ our Lord. Amen,’ occasionally elaborated into a Trinitarian form with the usual Roman collect ending ‘Who liveth and reigneth ...’ — an instance of the Roman temper of simplicity in such things.

After the sermon followed the dismissals of the catechumens and penitents and the intercessory ‘prayers of the faithful.’ These latter, a part of the synaxis but attended only by those about to attend the eucharist, had always formed a sort of intermediate section between the two rites when they were celebrated in sequence. The fusion of the two separate services in the fifth century did not destroy this special character of this part of the liturgy, though it brought changes of various kinds, due to the need for adapting the pre-Nicene tradition to the new purposes of a public worship. It was natural, too, that new items which it was desired to include somehow in the Shape of the Liturgy, but which had no obviously indicated place in the structure of the rite — e.g. the creed — should tend to be inserted at this point.

In the fifth century Christendom was markedly beginning to fall apart. The question of Byzantine centralization was dividing Syria and Egypt from the Balkan and Anatolian churches. The West was being parceled up between a number of barbarian tribal kingdoms, though the old Romanced populations carried on a good deal of the tradition of the fourth century underneath the political overlordship of the new masters, and the Western churches were now the mainstay of what remained of the civilized tradition. But the growing political divisions meant that the fifth century changes in the liturgy were carried out by churches no longer in the close contact with each other that alliance with the universal empire of the fourth century had ensured. The result is a growing diversity again (after the period of convergence in the fourth century) among the various liturgies, which-probably reached its height in the seventh-eighth century. After that the restoration of the Western empire by Charlemagne in a.d. 800 results in a general tendency towards uniformity in the West on the basis of the Roman rite, in the particular form in which the emperor had adopted this in his palace chapel. Despite a certain reaction against this ‘Romanism’ during the political confusion which followed Charlemagne’s death, most of the effects of his work were never undone in Western liturgy until the sixteenth century. In the East, the submersion of the Christian churches of Egypt and Syria under successive waves of Mohammedan conquest in the seventh-eighth centuries, eventually caused the Christians living as serfs under Islam to look towards Byzantium as in some sort the Christian stronghold of the East. Though their experience of Byzantine bureaucracy and Byzantine ecclesiastical politics had been so disastrous that they never forgot their bitterness against her sufficiently to enter again into communion with the Byzantine ‘orthodox’ patriarchs, yet Byzantium had at least the prestige of being the one free church of the East, and Byzantine ecclesiastical ways tended to spread among the dissidents in consequence.

The result of all this is a good deal of diversity in the arrangement of the items which belong to this ‘second stratum’ in the Shape of the Liturgy, and in the way in which various churches fitted them into the traditional outline which had come down everywhere unchanged from pre-Nicene times. Nevertheless one can distinguish certain groups in the table opposite (p. 475). I do not propose to go into all the diversities, some of which are unimportant. Others, however, have had a considerable effect upon the devotional ethos of the rites in which they are found.

The dismissals and prayers belong to the old synaxis. The latter were declining in popularity during the fifth century, arid being replaced either by newer forms of intercession like the litany, or by various ways of commemorating the living and the dead in the eucharist proper — a practice which I have called ‘the Names,’ to avoid begging certain questions connected with the particular custom known as the ‘recitation of the diptychs.’ Only in the Egyptian rites did the old ‘prayers of the faithful’ persist in something like their original form as well as position.

For the rest, the columns fall easily into two groups — those which have the ‘oblation’ by the people for themselves before the altar, comprising the Egyptian and Western rites; and those which have instead the ‘offertory procession’ of the deacons from the sacristy, in the form first fully described by Theodore of Mopsuestia. The primary example of these is the Byzantine rite; but the position of the offertory in the East Syrian and Jerusalem rites is somewhat obscure, though it is probable that they were both ‘procession’ rites, not ‘oblation’ rites, from at all events the fifth century.

A further interesting subdivision arises from the fact that all the Western rites seem to have stood together in the fifth century in placing a variable prayer before the Pax and the offertory, which I have called the ‘Prayer of the Day,’ of which all the Eastern rites (including the Egyptian) know nothing. The Western rites might in fact be placed in a single column in this table but for the awkwardness of shewing two facts. One is the curious position of the Pax (after the offertory prayer) in the Spanish and Gallican rites. This can hardly be its original position, but it was already placed there traditionally in Spain in the time of S. Isidore of Seville (c. a.d. 600), and there seems to be no evidence as to when or why it was moved from its (presumable) original position before the offertory, where it stood in all the pre-Nicene rites. The other point in which the Western rites vary among themselves is that in Spain and Gaul the recital of the ‘Names’ of the offerers is attached to the offertory, as early as can. 28 of Elvira (c. a.d. 305) in Spain; while at Rome and Milan, as in Egypt, it was inserted at an early point in the eucharistic prayer, and this apparently before c. a.d. 390.

In the Eastern rites, as in the Western, the offertory prayer naturally follows immediately upon the placing of the elements on the altar. This later insertion of an explicit offertory prayer links the offertory closely to the eucharistic prayer, but the Eastern rites have spoilt the connection by the insertion at this point of the creed, a late sixth century innovation, and the transference to this point of the Pax, originally the prelude to the offertory. The East Syrians seem never to have adopted the offertory prayer in the strict sense, retaining, I suppose, the primitive notion that the solemn placing of the elements upon the altar is an offering of them, needing no explicit verbal expression. It remains to discuss certain particular changes and insertions in the various rites, the reasons for them (where these can be discerned) and their consequences for the particular ethos and devotional convention of the rites in which they were made.

The litany form of prayer appears for the first time fully developed in the North Syrian rite of Ap. Const., viii. c. a.d. 370. It is interesting to note the exact forms in which it is found there. The dismissal of the catechumens begins by their being commanded by the deacon to kneel; he then proceeds to proclaim a series of petitions on behalf of them, to each of which the laity answer Kyrie eleison: ‘that He Who is good and loveth mankind will pitifully receive their prayers and entreaties’ (Kyrie eleison); ‘that He will reveal unto them the gospel of His Christ’ (Kyrie eleison); ‘that He will enlighten them and establish them with us’ (Kyrie eleison) — and so forth. These are prayers for the catechumens, in which they themselves take no part. After eighteen of these petitions, the catechumens are bidden to rise and then to pray for themselves: ‘Entreat for the peace of God through His Christ’; ‘Entreat that this day and all the days of your life be peaceful and sinless’; ‘that you make Christian ends,’ and so forth. Then they are bidden to bow for the bishop’s blessing, which he gives in the form of a longish prayer, and the deacon proclaims ‘Depart in peace, ye catechumens.’

There follow three more sets of dismissals on the same plan; for those possessed by evil spirits, those in the last stages of preparation for baptism and the penitents respectively. Each class is prayed over by the deacon and people in a scries of petitions with the response Kyrie eleison, and dismissed with the bishop’s blessing in the form of a prayer. The whole business seems very elaborate and can hardly have taken less than twenty minutes or so to perform. But the evidence of Chrysostom’s homilies preached at Antioch guarantees that the compiler has not imagined this system, but has on the whole kept faithfully to the Antiochene practice, though he has probably expanded it in some respects.

There follow the real ‘prayers of the faithful,’ intercessory petitions for the world at large proclaimed by the deacon, answered by the prostrate people with Kyrie eleison. But these petitions are slightly different from those said over the catechumens etc. in their construction: ‘For the peace and good order of the world and the holy churches let us pray; that the God of the universe may grant us His own everlasting peace that cannot be taken away and preserve us to pass all the days of our life in unmoved righteousness according to godliness.’ If we look back to the old intercessions (p. 42) we shall find that they consisted of 1. a subject given out by the deacon or celebrant, 2. the people’s prayer in silence, 3. a brief collect or prayer by the celebrant, summing up the people’s prayers. What seems to have happened here is that the celebrant’s collect after each pause for silent prayer has been slightly adapted and appended to the deacon’s bidding. ‘For the peace... let us pray’ is the old deacon’s bidding; ‘that etc’ (which has no parallel in the biddings over the catechumens) is the celebrant’s collect.

In form the change may not appear very great, but the effect is considerable. Under the old system the whole church did the substance of the praying, individually and in silence. The ‘liturgies’ of the deacon and celebrant only acted as a sort of ‘framework’ in what was a really corporate intercessory act. In the litany this has been altered. It has become a dialogue, between the deacon and the people, with the former very much predominant; and the celebrant has been eliminated. It is true that the people now have a vocal part, the Kyrie, but they are no longer the obvious active intercedes; they have become a sort of chorus. And the celebrant has been excluded altogether from the intercessions. It is true that in Ap. Const,, viii. the litany is followed by a prayer by the bishop. But if it be compared with the prayers over the catechumens and penitents which have just preceded the litany of intercession, it will be found that this prayer is not any summary or conclusion of the prayers ‘for all sorts and conditions of men’ which have been offered in the litany. It is a departure-blessing or dismissal of the faithful there present, a prayer for not with those who have been interceding, exactly comparable to the blessings of the catechumens, etc. before they leave the assembly. It marks the end of the synaxis, still an independent rite. Even if the eucharist is to follow, it may do so in another building or after an interval. But there is no justification in this case — or I would add in any other — for supposing that a prayer by the celebrant necessarily summed up or concluded the intercessory litany in the East. That consisted simply of the people’s response to the deacon’s petitions, which had absorbed the old celebrant’s part in the intercessions.

This curious evolution asks for some explanation beyond mere caprice, and it seems to have had an entirely practical origin. In Syria in the later fourth century there had been introduced the ‘sanctuary veil,’ a silk curtain cutting off the celebrant and the altar altogether from the sight of the congregation during the celebration of the eucharist.

To understand the real meaning and purpose of this innovation we must go back a little. We have already noted in S. Cyril’s Catecheses (Cf. p. 200) the first beginnings of the use of words like ‘awful’ or ‘terrifying,’ and the ‘language of fear’ generally, in reference to the consecrated sacrament. By the last quarter of the century this novel idea had taken a firm hold in Syrian devotion — it is notable, for instance, in Chrysostom’s sermons. Perhaps it found a specially congenial soil in Syria, where since time immemorial ‘the holy’ had also meant in some way ‘the dangerous’ (Cf. 2 Sam. vi. 7). It spread outside Syria northwards very soon. We find it, for instance, in Theodore (an Antiochene by training) at Mopsuestia (Cf. p. 283), who does not hesitate to say that the faithful ‘should be afraid to draw nigh unto the sacrament without a mediator and this is the priest who with his hand gives you the sacrament.’ We are evidently far in thought (but only a few years in time) from the days when the laity communicated themselves daily at dawn from the sacrament reserved in their own homes. It is a symptom of that decline — swift and sudden in the East, slower but steady in the West — in the understanding of the position of the laity as an ‘order’ in the church, a decline which begins in the fourth century. The word laikos ‘a layman’ in the East c. a.d. 300 still meant ‘one of the People (laos) of God,’ with all the rights and high duties and destinies that implied. By c. a.d. 450 it had almost come to mean ‘profane’ as opposed to ‘sacred.’ (There is required only one more step to reach the modern French meaning, e.g. in the phrase lois laiques where it means ‘anti-christian.’)

The veil which hid the sanctuary during the eucharist in the Syrian churches is the natural product of this frame of mind. ‘Liturgy’ is becoming the special function of the clergy alone, for their sacred character protects them in the ‘numinous’ presence of the sacrament, charged as it is with ‘terrifying’ power. The ‘profane’ laity have no such safeguard, and therefore the veil was introduced, to hide them from it rather than it from them. Perhaps the Old Testament precedent of the tabernacle veil had something to do with the innovation, but an origin in the same frame of mind rather than in deliberate imitation seems the truer explanation. And the earliest reference to the veil that I can find is in a homily of S. John Chrysostom preached at Antioch soon after a.d. 390: ‘When the sacrifice is borne forth (for the communion) and Christ the Victim and the Lord the Lamb, when thou hearest (the deacon proclaim) "Let us all entreat together ..." when you see the veil drawn aside — then bethink you that heaven is rent asunder from above and the angels are descending.’ There is no veil in Ap. Const. and it may not yet have been common outside Antioch. But if we are thinking of origins, I should be inclined to look behind Antioch to the church of the Anastasis at Jerusalem, where, as Etheria has told us (Cf. p. 438), the sacrament was consecrated, not exactly behind a veil, but still out of sight of the congregation, inside the cave of the Holy Sepulchre behind its great bronze screens. So far as the evidence goes, it was at Jerusalem that ‘the language of fear’ — which is at the very roots of this whole conception — first began to be used about the sacrament.

The atmosphere of ‘mystery’ and ‘awe’ which is the special ethos of the Byzantine rites seems to be very largely a product of the local churches of Syria in the fourth century. It is true that the veil in modern orthodox churches is only a relic of its former self, a mere door-curtain inside the central gates of a solid masonry screen, whose outer face is covered with the sacred ikons. The first occurrence of this further barrier between the laity and the consecrated sacrament seems to be in Justinian’s glorious rebuilding of the cathedral of the Holy Wisdom at Constantinople c. a.d. 570. It would appear, too, that in its main features (apart from the decoration with ikons, which may be a later development) this screen was originally nothing but a straightforward copy of the traditional back-scene of the Byzantine theatre with its three double doors. The idea was perhaps not so inappropriate as it may seem. The Byzantine rite had by this time taken on some of the characteristics of a drama.

What I am concerned to emphasise here is that the sixth century introduction of the solid screen at Constantinople did no more than confirm the great consequence of the introduction of the veil in Syria in the fourth century. This was the exclusion of the laity from the process of the liturgical action. When all has been said that is true — and very much is true — of the real spiritual participation of the orthodox laity at ail periods in the liturgical worship, it also remains true that the screen to a large extent forces upon the Eastern liturgies the character of two simultaneous services, the one proceeding outside the screen for the people, conducted chiefly by the deacon; the other — the real liturgical action — proceeding inside the screen conducted by the celebrant. Despite the general connection of the two and their spasmodic unification, and the function of the deacon who acts all the time as a connecting link, this duality is unmistakable at the actual performance of the liturgy in an orthodox church. And that character was originally given to it by the adoption of the veil and the hidden consecration in Syria during the fourth century. It is a quite different tradition of worship from our own, though we need not therefore condemn it or even criticize it. But we must grasp the essential difference between Eastern and Western eucharistic devotion, which begins in the fourth century — that while in the East the whole assumption and convention of the devotional tradition is that the people ought not to see the consecration, or indeed the progress of the liturgical action, in the West the devotional tradition assumes that they should see it. And when the new liturgical fashion for the ‘eastward position’ of the celebrant had for the first time made this difficult in the West, the new ceremony of the Elevation was deliberately introduced to shew them the sacrament.

The litany in the Eastern rite is more comprehensible in the light of all this. After the catechumens, etc. had retired the celebrant blessed the faithful at the end of their prayers as he had blessed the others, and so dismissed them in their turn. But if — as normally on Sundays — the eucharist was to follow, this final blessing of the faithful was not given. Instead the celebrant retired at once within the veil to prepare to celebrate, murmuring private prayers of deprecation for his own unworthiness (of the kind which now figure as the ‘prayers of the faithful’ in the Byzantine rites) leaving the intercessions to be conducted by the deacon outside the veil. It would be difficult, and in any case unedifying, to conduct the old ‘trialogue’ of deacon, people and celebrant through the curtain; it was much easier to allow the deacon to add the celebrant’s part in the intercessions to his own. Hence the litany.

Silent recitation — at least in great part — of the prayers at the eucharist would in any case have been likely to follow from this new separation of the celebrant and people, even if the psychological question of ‘reverence’ had never occurred to anyone. But it seems that in fact the latter was the determining cause of the introduction of the silent recitation of the eucharistic prayer, in the far East and the West at all events.

The main action of the eucharist was thus removed from the sight of the Eastern people. Except for the Great Entrance and the Communion all took place behind the veil or screen. It is not surprising that the ‘Great Entrance’ procession, when the sacrament was ‘carried to burial’ with solemn pomp, and its reappearance after an interval dramatically brought forth ‘resurrected’ at the moment of communion, became the twin focus of popular eucharistic devotion in the Greek churches.

Those who will may emphasise the ‘Eleusinian’ parallel thus produced in the Greek rite. For my own part I am clear that this interpretation of the eucharist was only built up by very gradual stages in the Greek churches and by successive independent changes in the presentation, not the contents, of the Greek liturgies, the prayers of which do not lend themselves very patiently to this interpretation. Some of the changes which ultimately had the most ‘Eleusinian’ effects began not in Greece at all but in Syria. Taking into account the late date at which the parallel — which can I, admit, be made to appear very striking — was finally developed, there can be little question of any direct imitation of hellenistic mysteries in the Byzantine rites. At the most all that could be suggested is a similar temper of thought underlying the Eleusinian mysteries and Greek eucharistic devotion. But we know too little about the former for any such parallel to be much more than an exercise of the imagination.

The Eastern people retained as their part in the liturgy listening to the lections (which the orthodox populations have always done with assiduity) and participation in some of the chants (though the admirable melodies of most of these were too difficult for the people and had to be left to the choir) — and the litany! It was natural this should be popular; it was the only devotion in the whole rite in which the laity as such now had any active part. From being used only at the intercessions which closed the synaxis it began to be repeated at other points in the rite, as an act of corporate prayer accompanying the liturgical action proceeding in mystery beyond the veil. It is now repeated no less than nine times in various forms, in whole or in part, during the Byzantine eucharist. With so many of the liturgical prayers said in silence, the litany forms the main substance of the people’s prayer.

There may be a certain evidence of liturgical decadence in this acceptance of the need to occupy the attention of the congregation with irrelevant devotions while the liturgical action — the eucharist proper — proceeds apart from them behind the screen. But even so, Westerns are hardly in a position to remark upon it. The Eastern litany is at least a corporate devotion provided by the church for the faithful, magnificently phrased and noble in its all-embracing charity. The Western ‘low mass,’ dialogued in an undertone between priest and server, is in a different way just as degenerate a representative of the old corporate worship of the eucharist. The faithful, it is true, can see the action and associate themselves continually with it in mind in a way that the Eastern layman cannot quite do. But the Western laity, unprovided with any corporate devotions whatever, are left with no active part in the rite at all. They listen and pray as individuals, adoring in their own hearts the Host elevated in silence, and then passively receive communion. All this throws the whole emphasis in Western lay devotion upon seeing, and on individual silent prayer. This question of ‘seeing’ is really at the basis not only of the difference of Eastern from Western eucharistic devotion, but of Western catholic and Western Protestant doctrinal disputes. Is what one sees elevated or ‘exposed’ — a significant word! — to be adored as such? Posed thus, apart from its context in the corporate offering, the question is distorted. But what caused it to be posed in this way in the sixteenth century, and made the reality of the Body and Blood of Christ a center of controversy in the West as it never had been in the East, was precisely the growth of low mass as the normal presentation of the eucharist to the laity during the mediaeval period.

We see, too, now why the litany never proved nearly so popular in the West as in the East. Though it was introduced at some time or another into most Western rites — I see no evidences that it was ever used in Africa — it disappeared from them again often without trace, because the people felt no need of it. It was the Eastern laity’s substitute for seeing the action of the liturgy, their way of associating themselves with it beyond the screen. The particular conditions which made it so popular in the East simply did not exist in the West, where the people found other substitutes in sight and private prayer for their old active participation in the rite.

The introduction of the creed into the I liturgy has a curious history. Its original usage was at baptism. From the "earliest days repentance and the acceptance of the belief of the church was the condition sine qua non of baptism into the Body of Christ, and formal interrogation as to both was made of converts before they received tithe sacraments. A statement of belief that ‘Jesus is Messiah’ with all that this implied might be accounted sufficient in Jewish circles, with their background of unwavering monotheism. But more was rapidly found necessary among the gentiles, to furnish security that the convert was not simply accepting ‘the Kyrios Jesus’ as one more ‘Saviour’ among hiss ‘gods many and lords many.’ The baptismal creed was elaborated as a series of three questions dealing respectively with the three Persons of the Holy Trinity, and clear traces of it in this short form are to be found in thee first half of the second century. The prevalence of gnosticism with its denials of the goodness of creation and the reality of our Lord’s Manhood brought further elaboration in the later second century — the affirmations that ‘God the Father’ is "Maker of heaven and earth’ (and therefore that creation is essentially good as the act of a good God); that Jesus Christ is not only ‘His only Son’ and ‘our Lord,’ but was truly conceived and born of a human mother, the Virgin Mary, and truly ‘suffered’ at a particular point in history ‘under (i.e. in the governorship of) Pontius Pilate’ and ‘died’ as all men die, and was ‘buried’ as a dead body (and was not spirited away into heaven. from the Cross or before the crucifixion, as the gnostics taught); and further that ‘the Holy Spirit’ is ‘in the Holy Church’ (alone, not in self-constituted gnostic cliques). We find it in this form in Hippolytus’ account of baptism, as a threefold question and answer, in a text which is the obvious j parent of our ‘Apostles’ Creed.’

The Council of Nicaea in a.d. 325 carried the use of the creed a stage further. It was no longer to be only a test of belief for those entering the church from outside. Since misbelief had shewn itself to be prevalent in the East not only among those who had been "baptized but amongst bishops and clergy, the creed was to be made a nest for those already within the church, by solemn affirmation of which they might prove that they believed what the church had always believed and! no t some new private invention of their own. And since the old formulae, however well they might serve to distinguish a pagan or a jew from a Christian, were too imprecise to distinguish an Arian from an orthodox Christian, the Council drew up a new creed, that which in an elaborated form we know as the ‘Nicene Creed.’ The basis appears to have been the old baptismal creed of Jerusalem, but the council added to the second section dealing with our Lord Jesus Christ a carefully worded formula — ‘God of God, Light of Light, very God of very God begotten not made, of the being of the Father, of one substance with the Father,’ which no Arian could conscientiously recite. In acting thus the Council was acting in precisely the same way as the church of the second century in adding the anti-gnostic clauses, and indeed as the apostles had acted in requiring the original affirmation that ‘Jesus is Messiah,’ which no unconverted member of the old Israel would make.

The precise stages by which the Nicene Creed as drawn up by the Council became our present ‘oecumenical’ or ‘Niceno-Constantinopolitan’ Creed are obscure. What is certain is that the Council did not draw it up with any intention of inserting it into the liturgy in any connection, and that it did not replace the older local creeds at baptisms, even in the East, for a considerable time. In the West the old Roman creed which we call the Apostles’ has everywhere persisted to this day as the test of a catechumen’s faith at baptism. The Nicene Creed was a theological statement of the church’s faith for Christians, not a test for converts from paganism.

In the monophysite troubles of the fifth century which followed upon the Council of Chalcedon (a.d. 451) it became the policy of the monophysite or federalist party to cry up the Council of Nicaea in order to slight ‘the emperor’s Council’ of Chalcedon, which they rejected. With this end in view the monophysite patriarch of Antioch, Peter ‘the Fuller’ in a.d. 473 instituted the custom of publicly reciting the Nicene Creed at every offering of the liturgy, as an ostentatious act of deference towards the venerable Council of Nicaea, whose teachings he declared that the Chalcedonians had abandoned. In A.D. 511, the patriarch Macedonius II of Constantinople — a pious but not very wise eunuch — was banished and deposed by the monophysite emperor Anastasius, after a series of diplomatic maneuvers which has few equals for unsavouriness even in the annals of Levantine Christianity. Macedonius’ intruded successor, Timothy — a man who appears to have had as little real concern for Nicene theology as for the Ten Commandments — at once introduced the monophysite practice of reciting the Nicene Creed into the liturgy of Constantinople, in order to secure the political support of the monophysite emperor and the federalist party generally. When by the vicissitudes of political fortune the orthodox once more secured control of the see of Constantinople, they dared not incur the odium of seeming to attack the memory of Nicaea by discontinuing this use of the creed; and so this originally heretical practice became a permanent feature of the Byzantine liturgy.

The West held aloof for a while, but the third Council of Toledo (can. 2) in a.d. 589 directed that ‘For the fortifying of our people’s recent conversion’ from Arianism the creed should be recited ‘after the fashion of the Eastern fathers’ by all in a loud voice. But this Spanish Council placed this recitation after the fraction ‘that first the people may confess the belief they hold, and then with their hearts purified by faith’ proceed to their communion. Its adoption among the Goths in Spain thus put it to its original purpose as a test for Arians, but in a new way, by making its recitation a preliminary to communion. In this unusual position it remains in the Mozarabic rite. Spanish Catholicism was always apt to make use of its belief as a weapon, witness the ‘damnatory clauses’ of another Spanish document, the so-called ‘Athanasian Creed.’ It was in Spain also that the Filioque clause was first added to the Nicene Creed as an anti-Arian declaration, which subsequently caused so much unnecessary trouble between the West and the East.

In Gaul the emperor Charlemagne seems to have been the first to introduce the singing of the creed, in the liturgy of his palace chapel at Aix in a.d. 798. Some other churches of his dominions did not adopt it until almost a century later, but it spread generally in Frankish churches fairly quickly. Some Frankish monks at Jerusalem got into trouble for singing it with the Filioque as early as a.d. 806, and defended themselves with the plea that they had heard it ‘sung thus in the West in the emperor’s chapel.’

Charlemagne thus used the Spanish text of the creed, but he did not place it at the Spanish position after the fraction, but where we now recite it, immediately after the gospel. There seems to be no doubt that this was a usage which had been growing up in the Italian churches outside Rome. It stood in this position in the rite of Benevento in the eighth century, and there is some evidence that the same custom had been introduced at Aquileia in N. Italy by its bishop Paulinus (a.d. 786-802). Rome, perhaps from mere conservatism, or perhaps misliking the heretical origin of the custom, long held out against the innovation. The recitation of the creed at the eucharist was first adopted by Pope Benedict VIII in the year a.d. 1014, under strong pressure from the Emperor Henry II, who was shocked when visiting Rome to find that it had no place in the Roman rite as it had in that of his own chapel. Even then Rome adopted it somewhat half-heartedly. It never became there, as in the East, an invariable element of the rite, but was reserved for Sundays and the greater feasts, as an appropriate expansion offering opportunities for singing. In later times there has been added the recitation of the creed at the eucharist on the minor feasts of those saints who are venerated as ‘Doctors of the Church,’ who by their writings have expounded and defended the faith which the creed sets out. Once more we can trace the repugnance to the Roman liturgical instinct of all additions to the rite which play no clear logical part in the performance of the eucharistic action, and so may confuse the bare simplicity of its outline, even while adorning it.

This prayer is peculiar to the Western rites. It seems to have stood at the same point in all of them in the fifth century, viz. after the dismissals which closed the synaxis and before the kiss of peace which formed the ancient opening of the eucharist. It thus formed a new opening prayer to the eucharist proper. It varied with the day, and its introduction is probably one of the earliest examples of that special influence of the calendar on the prayers of the eucharistic rite which is a peculiar feature of the Western liturgies as a group. The simplest thing is to give some examples of this prayer in the various rites, beginning with that in which it is most fully developed and has most completely maintained its function, the Spanish Mozarabic rite.

In the Spanish books this prayer is always constructed in two parts, the one addressed to the congregation, the other directly addressed to God, known respectively as the missa and alia — ‘the mass’ and ‘the other’ (prayer). Here is the ninth century Mozarabic prayer ‘of the day’ for Tuesday in Holy Week:

Missa: ‘Offering the living sacrifice to our most loving God and Redeemer, we are bound, dearly beloved brethren, both to entreat Him by our prayers and do penance by our tears: for His holy Pascha draws near and the celebration of His passion is at hand, when by the penalty of the torment laid upon Him He burst the gates of hell. Let us serve Him by fasting and worship Him by contrition of heart, seeking of Him that He will through abstinence cleanse our flesh burdened with sins and rouse our dull mind to love Him by the approaching celebration of His death.

Alia: ‘O Christ our Saviour, God, at the approach of Whose passion we rejoice, and by the yearly return of the celebration of Whose resurrection we are raised up: do Thou cleanse our flesh brought low with fasting from the weight of our sins. Do Thou sanctify the soul that has earnestly desired Thee: grant light unto the eyes: give cleanness to body and soul: that worthily adorned (vestiti) with all virtues we may be found meet to behold the glory of Thy passion.’

Though the Mozarabic terms are missa and alia, this is an example of the old Western praefatio and oratio structure of which we have already spoken, the two parts forming a single prayer. After the praefatio there was originally a pause for silent prayer, followed by the celebrant’s oratio. We have had an example of the same structure in the Roman intercessory prayers, with the celebrant’s bidding ‘Let us pray, beloved brethren, for.. .’ (followed by the deacon’s command to kneel at the great intercessions, and perhaps on other occasions in penitential seasons) and then after the people’s silent prayer, the collect (Cf.p.42). Another survival of the same thing in the Roman rite is the celebrant’s address before the Lord’s prayer after the canon, ‘Let us pray: Instructed by saving precepts and taught by divine example we make bold to say: Our Father. . ..’ In this case the Lord’s prayer itself takes the place of the pause for silent prayer, and the celebrant concludes with a collect which is now said inaudibly in the Roman rite (except on Good Friday at the communion of the Pre-sanctified) but is still always recited aloud at Milan. Other Roman survivals of the full praefatio are to be found before the collects in ordination masses. Indeed, it has not entirely disappeared before any Roman collect, for the celebrant always ‘prefaces’ his ‘prayer’ (oratio), addressed to God, with Oremus, ‘Let us pray,’ addressed to the people. The Eastern rites have no such address before their prayers. It is very typical of the different genius of the two Western liturgical types, Italian and Franco-Spanish, that starting from the same sort of formula of a single sentence or so, the one should tend to cut it down always to the same single word, and the other should expand it to a paragraph or more (some Mozarabic praefationes are fifty or sixty lines long) and vary it on every occasion that it is used.

There are sufficient indications that throughout the West all the prayers of the liturgy except the eucharistic prayer were at one time constructed in this way, with an address to the people followed by the prayer proper. But by the time our oldest extant liturgical MSS. were written the system was in full decay, the address being often reduced to a few words, or more usually omitted altogether. The cumbersomeness, and also the somewhat offensive clericalism, of prefixing an exhortation to the people by the priest every time prayer was to be offered was too much for the tradition. And even Spanish fecundity of liturgical expression boggled at the task of finding a sufficient number of different ‘prefaces’ for all the variable prayers of this most mutable rite. The missa and alia, however, in the Mozarabic rite retained the full form and even expanded it considerably, for a particular reason. There was no collect or other prayer in the Mozarabic rite before the lections until the tenth century or so. Thus the missa and alia together formed the first prayer of the day, and had the function of ‘striking the keynote’ as it were of the special liturgical character of the mass. When the variable ‘collect’ before the lections was introduced into the Spanish rite, it more or less duplicated this function; but by then the missa and alia were too strongly entrenched in Mozarabic tradition to be attenuated. At Rome the ‘prayer of the day’ disappeared, but it was the ‘collect’ not the ‘prayer of the day’ which tended to be eliminated in Spain, being altogether omitted on all fast days. Mozarabic masses were cited by the first words of this prayer (whence the name missa?) just as Roman masses were and are cited by the first words of the introit, as a convenient way of referring to the mass of different occasions and days (e.g. Requiem, Laetare, Quasimodo, etc.).

In Gaul we find the same arrangement of praefatio and oratio at the same point of the rite. But here the Roman ‘collect’ before the lections was introduced much earlier than in Spain (sixth-seventh century?) and in the later Gallican books is already tending to oust the praefatio and oratio from their original function of emphasizing the particular point of the liturgy of the day. Originally the Gallican ‘collect’ before the lections appears to have had the character of a mere preparatory prayer, leaving the reference to the saint or the day to the prayer ‘of the day’ after the gospel. The following, from the mass of S. Germanus of Autun in the oldest Gallican collection extant, the Masses of Mone, will make the difference plain:

Collect (before the lections): ‘O pitiful and pitying Lord, Who if Thou didst repay us according to our deserts, wouldst find nothing worthy of Thy forgiveness; multiply upon us Thy mercy that where sin hath abounded, the grace of forgiveness may yet more abound. Through ...

Praefatio (after the gospel): ‘With one accord, my dear brethren, let us entreat the Lord that this our festival begun by the merits of our blessed father the bishop Germanus may by his intercession bring peace to his people, increase their faith, give purity of heart, gird their loins and open unto them the portal of salvation. Through ...

Oratio ante nomina. ‘Hear us, O Lord holy, Father Almighty, everlasting God, and by the merits and prayers of Thy holy pontiff Bishop Germanus, keep this Thy people in Thy pity, preserve them by Thy favour, and save them by Thy love. Through ....’

At Milan the prayer ‘of the day’ is known as the ‘prayer over the corporal’ (oratio super sindonem) i.e., the first prayer said after the cloth has been spread by the deacon upon the altar, which as we have seen (Cf. p. 104) was the first preparation made for the celebration of the eucharist proper. It is preceded by The Lord be with you,’ ‘And with thy spirit,’ and ‘Let us pray’ — precisely like the collect before the lections, from which in the Milanese rite it is indistinguishable in function by its contents. Indeed a few prayers which are employed in one Milanese MS. as collects proper are exchanged by others with the corresponding super sindonem prayers, without the mistake being detectable from the contents of the prayers.

In the Roman rite there is no longer a prayer ‘of the day.’ But before the offertory the celebrant still turns to the people for ‘The Lord be with you,’ ‘And with thy spirit,’ and turns back saying ‘Let us pray’ — but no prayer follows. Something has dropped out of the rite, and the close analogy of Milan suggests that it is a super sindonem prayer. Nor perhaps are we altogether without information as to the actual ‘prayers of the day’ used on some of the days of the liturgical year at Rome in the fifth-sixth century. Liturgists have long been puzzled to account for the fact that while the masses of the Roman Gregorian Sacramentary have only a single collect before the lections, the pre-Gregorian Gelasian Sacramentary usually gives two. A certain number of these supplementary Gelasian collects reappear in the Milanese rite as orationes super sindonem. I suggest that when the prayer ‘of the day’ was abolished at Rome (was it by S. Gregory?) some Italian church south of Rome did not at once follow suit, and retained the super sindonem prayers. Our unique copy of the Gelasim Sacramentary, though it reproduces the substance of a pre-Gregorian Italian book, was made in France c. a.d. 700. It was thus written a century or more after the Gregorian reform (c. a.d. 595) and with full knowledge of the changes introduced by S. Gregory, to which in many important details it has been accommodated (e.g. it incorporates all the changes he had made in the text of the canon). But it descends, so far as its ‘propers’ are concerned, not from a sacramentary used in the city of Rome itself, but from an Italian book from the country south of Rome (? Capua), as is proved by its calendar. I suggest that this South Italian book retained the super sindonem prayers, which the scribe of our Gelasianum MS. has preserved, merely omitting their headings to bring the copy he was making into line with the current Roman and Frankish use.

We can, I think, understand the disuse of the prayer ‘of the day’ in the Roman rite. Once the variable ‘collect’ before the lections had made good its footing in the rite, it anticipated the function of the prayer ‘of the day’ after the lections. The first prayer thus ‘struck the key-note’ of the day at a more appropriate point in the rite than did the second, once the lections of the synaxis had come to be thoroughly fused with the eucharist proper as parts of a single whole. And so, finding itself with what were virtually two ‘collects,’ one before and one after the lections, both fulfilling the same function, the Roman church dropped the ‘prayer of the day’ at some time in the sixth century in favour of the ‘collect’ before the lections; though the latter was a custom imported from Egypt in the course of the fifth century, while the prayer ‘of the day’ was an element in the Roman rite which it shared with the other Western churches.

We have insufficient evidence about the African rite to be sure whether it contained a prayer ‘of the day,’ though there are texts which might reasonably be conjectured to refer to it.

The interest of this prayer ‘of the day’ is twofold. First, it is a feature which is common to all the Western rites and missing from all the Eastern ones. It thus gives an indication that the Western rites under their later divergence originally form a real group, going back to a common type. Secondly, from its character and position its introduction must go back to the period before the synaxis and eucharist were properly fused, but after the formation of the liturgical year — say round about a.d. 420-30. Only at that time could it have been felt necessary to insert a prayer specially intended to bring the fixed prayers of the eucharist proper into direct relation to the lections that had just been read, and to the day in the liturgical calendar. Its institution is thus probably the earliest effect of the calendar on the prayers of the eucharist, which became so marked a feature of all Western rites in the fifth century and after.

We have seen that the offertory procession at Mopsuestia in Theodore’s time advanced from the sacristy to the altar in dead silence, a point on which Theodore lays special emphasis (Cf. p. 283); and there is no mention of music or singing at this point of the rite in Narsai. It is interesting to find that the Western oblation by the people before the altar appears also to have been originally performed in silence. The interest of the pre-Nicene church both East and West is concentrated on the action of offering. No need was felt to ‘cover’ this, as it were, with music. The first we hear of an offertory chant is from S. Augustine in Africa, who notes in his Retractations the introduction in his own days at Carthage of’the custom of reciting (dicerentur) at the altar hymns taken from the book of psalms both before the oblation and while that which had been offered was being distributed to the people,’ and how he himself had been obliged to write a pamphlet in defense of die innovation.

So far as can be made out from the obscure and scanty evidence the original form of this psalmody was what the ancients called ‘responsorial,’ i.e., a solo singer sang the verses of a psalm to an elaborate setting, the people and choir joining in with a chorus or refrain between each verse — the ‘antiphon.’ At Rome, when the offertory and communion psalms were adopted, a plain psalm chant sung by the people seems to have been adopted for the verses, the ‘antiphona’ melody being more elaborated and left to the choir. When the people’s oblation gradually fell into disuse on normal occasions (as lay communions grew more infrequent) less music was required to ‘cover’ the offertory ceremony; and so the psalm verses were cut down until by degrees they vanished altogether (except at requiems), leaving only the elaborate melody of the antiphon to be rendered once by the choir as a sort of ‘anthem’ at the offertory. The same thing happened with the communion psalm. But two or three psalm ‘verses’ are still found on occasion attached to the antiphon in Roman choir books of the eleventh century. We do not know when the Roman church adopted the African custom of singing psalms at the offertory and communion, in addition to the pre-Nicene chants between the lections and its own early fifth century innovation of a psalm-chant during the processional entry. But a careful study of the texts of the offertories and communions in the Gregorian antiphonary suggests that they are a later development than the introit psalms. Not only are there few (if any) survivals of the pre-Vulgate text of the scriptures in these chants, of a kind which are not infrequent in the graduals and tracts and found occasionally in the introits; but they are usually chosen without close connection with the introit (which often has a connection of thought with the gradual). On the other hand, offertory and communion often seem to have a connection of thought between themselves. Perhaps a simultaneous adoption at Rome later in the fifth century than the introit would satisfy all the known facts.

The Western rites thus equipped themselves with offertory chants independently of and before those of the East. There does in fact seem to have been much more general interest taken in church music in the West than in the East from the fourth century onwards. There was singing in the Eastern liturgies, at all events in the synaxis and (after its adoption) at the sanctus of the eucharist. The Eastern rites would have been untrue to the primordial origins of the eucharist in the chaburah supper with its psalm-singing if they had excluded singing altogether. But if one looks at an Eastern exposition of the liturgy earlier than the seventh or eighth century, whether it be Cyril of Jerusalem or Theodore or Narsai, one finds that when music is mentioned it is passed over as something incidental, which excites no interest. In the West there is a series of writers beginning with Augustine who discuss with evident appreciation the part of church music in worship, its legitimacy, its appropriateness and emotional effects, in a way which so far as I know is unparalleled in the East at this date. And whereas when the Eastern writers wish to dilate on the impressiveness of the eucharistic rite their emphasis is regularly on what strikes the eye — on the ceremonial and the vestments — comparable Western writings like S. Isidore de Offisis and pseudo-Germanus lay their emphasis rather on the splendour of what is heard — the church music; and they evidently ascribe the same sort of emotional effect to this as is made on the Easterns by the ceremonial.

There is here not much more than a difference of psychology, so far as the early centuries are concerned. The Easterns developed a church music of a very high order. The researches of Herr Egon Wellescz and Professor Tillyard are teaching us that Byzantine church music of the golden age (much of which has a Syrian origin) was equal to the best that the West could produce. And the Westerns developed a ceremonial, stately enough in its own way though it never attained to anything like the dramatic quality found in the Eastern rites; and in Gaul (and perhaps during the middle ages generally) Western ceremonialists were apt to mistake mere fussiness and elaboration for dignity. But that the popular emotional interest in the East and West varied between ceremonial and music in the way described seems clear. This had some effect on the later liturgical history of the two halves of Christendom. It was the special perfection and completeness of the Roman chant which as much as anything else spread the Roman rite in the West from the eighth century onwards, for the chant fitted the rite and it was difficult to adopt one without the other. But it is the spread of Byzantine ceremonies (e.g. the ‘prothesis’ or ceremonial preparation of the elements before the synaxis, and the ‘great entrance’) which has so largely Byzantinised the rites even of the dissidents in the East.

The Western appreciation of and interest in the music of worship has survived even the triumph of the puritan ideal among the churches of the Reformation, except among the most austerely consistent of the sects. This is true not only e.g. of the Anglican Cathedral tradition,’ but among Prussian Calvinists whose grim worship still admits their lovely chorales. The point is that oriental puritans admit no such illogicality. Islam has neither instrumental nor choral music in its corporate worship. As a mohammedan mallaum once shrewdly remarked to me of a Wesleyan mission — ‘They will have beautiful sounds but not beautiful sights or odours like you in their worship. Yet the sounds are more distracting from true prayer than the sights or odours would be, which is why we true believers admit only words.’ That is the puritan theory of worship in a nutshell — to ‘admit only words.’ The Western interest in ‘church song’ which begins in the fourth century with Ambrose and Augustine has certainly shewn itself very strong to overcome this instinct of puritanism in any department of worship. It is curious that it has nowhere (I think) been strong enough to retain among protestants the old recitative or intonation of lections and prayers to a very simple chant as in the synagogue and the primitive Christian church — the one and only sphere in which Islamic custom has preserved music in its liturgy.

We have seen that none of the pre-Nicene rites contain any offertory prayer at all. The interest is concentrated upon the action, and the setting of the bread and wine upon the altar in and by itself constitutes the offering of them to God. The addition of an explicit commendation of them to God is an innovation of what I have called the period of the ‘second stratum,’ the fifth-eighth centuries. It is an indication that the period when the eucharist is recognised as primarily an action, in which every member of the church has an active part, is passing into the later idea of the eucharist as primarily something ‘said’ by the clergy on behalf of the church, though it was centuries before this idea took complete control of the presentation of the liturgy.

There is still no offertory prayer in Sarapion; nor is there any such prayer in Ap. Const., viii, thirty or forty years later in Syria. There is no means of telling how old the offertory prayer found in the ninth century text of the liturgy of S. Basil may be, but it is likely to be as ancient as any used in the East and is in itself so fine a prayer as to be worth citing as a representative of the later Eastern prayers:

‘O Lord our God Who didst make us and bring us into this life, and show us the ways unto salvation, and grant us the grace of the revelation of heavenly mysteries: Thou art He Who did set us in this ministry in the power of Thy Holy Spirit. Be graciously pleased, O Lord, that we should be ministers (diakonous) of Thy New Covenant, officiants (liturgisers, leitourgous) of Thy holy mysteries. Receive us as we draw near unto Thy holy altar in the multitude of Thy mercy that we may be made worthy to offer unto Thee this reasonable and unbloody sacrifice on behalf of our own sins and the ignorance of the people. Receive it upon Thy holy and heavenly and spiritual altar for a savour, of sweetness, and send down in return upon us the grace of Thy Holy Spirit. Look upon us, O God, and behold this our worship, and accept it as Thou didst accept the gifts of Abel, the sacrifices of Noah, the whole-burnt-offerings of Abraham, the priestly offerings of Moses and Aaron, the peace-offerings of Samuel; as Thou didst accept from Thy holy apostles this true worship, so accept also from the hands of us sinners these gifts in Thy goodness, O Lord, that being found worthy to liturgise blamelessly at Thy holy altar we may receive the reward of faithful and wise stewards in the day of Thy righteous repayment, through the mercies of Thy only-begotten Son with Whom Thou art blessed with Thine all-holy and good and life-giving Spirit, now and for ever and for ages of ages. Amen.’

The earliest suggestion of such a prayer in Christian literature is, as we have said, in the letter of Pope Innocent I to Decentius (c. a.d. 415), but we have no evidence when the Roman prayers first assumed their present form, of which the following are specimens taken almost at random:

For the Epiphany: ‘We beseech Thee, O Lord, graciously to behold the gifts of Thy church: wherein is set forth no longer gold and frankincense and myrrh, but what by those gifts is declared and sacrificed and received, even Jesus Christ Thy Son our Lord...’ For the second Sunday after Epiphany: ‘Sanctify, O Lord, our offered gifts: and cleanse us from the stains of our sins; Through...’ For Low Sunday: ‘Receive, we pray, O Lord, the gifts of Thy jubilant church, and since Thou hast given her reason for such mighty joy, grant her also the fruit of endless bliss. Through...’ For the fifth Sunday after Pentecost (fourth after Trinity): ‘Be gracious, O Lord, unto our supplications and mercifully receive these oblations of Thy servants and handmaids; that what each has offered to the honour of Thy Name, may avail for the salvation of all; Through...’

These set forth with simplicity the spirit of the people’s oblation, brought into contact now with the offerings of the wise kings, now with the thrill of the Easter joy, and in the ‘green’ seasons with the endless desire of the soul for purity and salvation.

The offertory prayers of the other Western rites are rather less directly expressed. Here for instance is the Mozarabic post nomina or offertory prayer for Easter Day: ‘Having listened to the names of those who offer, we pray Thee, Lord of love, to deign to be present to us at our prayer, to be found when Thou art sought, to open at our knocking. Write the names of the offerers in the heavenly book, shew forth Thy promise in the holy, Thy mercy in the lost. And because the prayer of our infirmity is weak, and we know not what to ask, we call to the aid of our own prayers the patriarchs taken into the heavenly company, the prophets filled with the divine Spirit, the martyrs crowned with the flowers of their confession, the apostles chosen for the office of preaching. Through whom we pray to Thee, our Lord, that all who are terrified by fear, afflicted by want, vexed by trials, laid low by sickness, bound captive by sufferings, may be released by the presence of Thy resurrection. Be graciously mindful also of the spirits of them that sleep (pausantium), that the outstretched pardon of their offences may allow them to attain to the bosom of the patriarchs, by the help of Thy mercy Who livest…’

The custom of reading out ‘the Names’ between the oblation and the offertory prayer in the Spanish church, and also the adoption of the oriental fashion of the diptychs have done a good deal to confuse the tenor of most of the Spanish offertory prayers. But even making allowance for this, there is usually a lack of simplicity about them and a striving after effect which results in turgid language; here, for instance, the allusion to Easter as ‘the presence of Thy resurrection’ releasing sufferers is clumsily made. One reason at least why the Roman rite was so largely adopted in the West without compulsion and by the gradual acceptance of so many local churches in the seventh-tenth centuries lies precisely in this, that on the whole it was a simpler and more expressive rite. The old local rites were redolent of the soil on which they arose, and rightly dear to those who used them from ancestral tradition. But rite for rite and prayer for prayer the Roman was apt to be both more practical and better thought out; and those who compared them carefully could hardly fail to notice it. Hence the growing voluntary adoption of Roman prayers and pieces and chants, and ultimately of the Roman Shape of the Liturgy as a whole, which is so marked a feature of liturgical history in the territories of the Gallican and Mozarabic rites during the seventh and eighth centuries, when the Popes were in no position to bring pressure to bear on anyone to adopt their rite.

To complete our survey: the Milanese offertory prayers, though by no means identical with the Roman series, are cast in the same mould, and need not be illustrated. The Gallican ones are usually similar to the Mozarabic. The following from the Missale Gothicum for Easter Day will serve for comparison: ‘Receive, we beseech Thee, O Lord, the Victim (hostia) of propitiation and praise and be pleased to accept these oblations of Thy servants and handmaids which we offer at the Resurrection of our Lord Jesus Christ according to the flesh. And grant also by the intercessions of Thy saints unto our dear ones who sleep in Christ refreshment in the land of the living: Through ...’ Here again the reading of ‘the Names’ of the departed and the saints immediately before has produced the incongruous addition of the last clause; though the undying French devotion to the memory of their dead, their cari nostri, which on the 2nd of November can still always bridge the great gulf between the French clerical and anticlerical, is something very near the heart of French religion in every age.

The intercessory ‘prayers of the faithful’ at the synaxis, like the petitions of the later litany which replaced them in the East, were general prayers — i.e., they spoke of classes of people, catechumens, penitents, travelers, pagans and so forth, without specifying individuals. The congregation were no doubt expected to particularize silently those in whom each was personally interested during the pause between the bidding and the collect. The only names publicly mentioned seem to have been those of the Roman emperor and the local bishop. But while this public intercession ‘by categories’ sufficed at the synaxis, the eucharist even in pre-Nicene times was felt to require something more personal, as the domestic gathering of the household of God.

It may be that the need for particularization was first felt at that peculiarly personal occasion, the eucharist offered for a departed Christian, when S. Paul’s teaching that the eucharist is always an anticipation of the judgement of God (1 Cor. 11:29-32) takes on a special poignancy. At all events, the earliest mention of the naming of an individual in the prayers of the eucharist proper, in the first epistle of S. Cyprian of Carthage, occurs in this connection. It deals with the awkward case of a bishop lately dead, who had deliberately violated a rule made by a recent African Council against the inconvenient practice of appointing clergymen as executors. Cyprian decides in accordance with the Council’s ruling that ‘there shall be no oblation on his behalf (at the offertory) nor shall the sacrifice be offered for his repose, for he does not deserve to be named in the prayer of bishops who has sought to distract the bishops and ministers from (the service of) the altar.’ Thus in Africa c. a.d. 240 it was already customary to name individual dead persons in the course of the prex, the eucharistic prayer, at all events at funerals and requiems. (Cyprian is not legislating for the deceased’s own church, where the actual funeral would take place, but for Carthage and other churches where a eucharist would customarily have been ‘offered for his repose’). It may be an accident, but Cyprian appears never to mention any ‘naming’ of living individuals at any point of the rite. S. Augustine a century and a half later has likewise no mention of the ‘naming’ of living individuals in the African rite, but his evidence as to the ‘naming’ of the dead is difficult to interpret. What he does make clear is that by his time the Jerusalem practice of ‘naming’ certain martyrs in the course of the eucharistic prayer had been adopted in Africa.

We have already noted (Cf. p. 164) that in Sarapion’s eucharistic prayer there is to be a pause for ‘the reading out (hypobole) of the names’ of the departed only. Likewise in Cyril’s account of the Jerusalem rite particular dead persons are named in the intercessions which follow the consecration, because of ‘the special assistance of their souls for whom prayer is made in the presence of the holy and most awful sacrifice.’ From the defense of the practice which Cyril thinks it right to make, one would suppose that this naming of individual souls in the eucharistic prayer was a fairly recent innovation at Jerusalem, and had been causing some discussion there. In Sarapion’s rite likewise, the ‘naming’ of the dead appears to be a fairly recent interpolation, with no connection with what precedes and follows.

So much for the early evidence for the naming of the dead at the eucharist. Now as concerns the naming of the living. The earliest evidence of this comes from Spain. The Council of Elvira (c. a.d. 305) in its 29th canon forbids the names of those possessed by evil spirits ‘to be recited at the altar with the oblation.’ Canon 28 prohibits an abuse which had grown up by which persons under excommunication — probably those who for social reasons had made some excessive compromise with pagan conventions — were allowed to offer their prosphora and have their names read out with the rest, provided they did not actually make their communion. All this would suggest that this ‘naming’ of the living in the Spanish rite was practically a roll-call of the faithful, and took place as each made their oblation or perhaps all together immediately afterwards. We can see now why the Spanish offertory prayers are called ‘the prayer ad nomina’ (‘at the names’) and why they take the form they do. In a small church where the members were well known to one another the omission of a name week by week would leave a stigma, and perhaps that is the origin and purpose of the custom. The ‘Names’ are those of the communicants (or ‘offerers’ as the ancient church thought of them) of that particular eucharist. Some of the later Mozarabic prayers are explicit that they are the names offerentium et pausantium ‘of the (living) offerers and the departed.’ It is possible that this was already so in pre-Nicene times, the relatives or representatives of the dead offering in ‘the name of those departed from that church in its peace and communion, a touching illustration of the vividness of belief in the communion of saints and the unity in Christ of all Christians living and dead. But though the early Spanish evidence does not contradict such an idea, it does not explicitly support it. Early practice in Africa and Spain was evidently not the same. Cyprian’s ‘naming’ of the dead is in the course of the eucharistic prayer. The Spanish ‘naming’ of the (living) ‘offerers’ is before it begins.

Before turning our attention to the East it will be as well to take here the earliest Italian evidence on the subject, though it is only at the end of the fourth and early in the fifth century that any is available. S. Ambrose at Milan tells us that ‘prayers are asked for kings, for the people and the others’ at an early point in the eucharistic prayer itself. We shall find that another N. Italian prayer of about the same date seems to have had the same arrangement. It is also the point which seems to be implied in Innocent I’s description of the Roman rite c. a.d. 415: ‘Your own wisdom will shew how superfluous it is to pronounce the name of a man whose oblation you have not yet offered to God (?by the offertory prayer). ... So, one should first commend the offerings and afterwards name those who have made them. One should name them during the divine mysteries and not in the part of the rite which precedes, so that the mysteries themselves lead up to the prayers to be offered.’ Whether the offerings are here ‘commended’ to God by a detached offertory prayer proper, or whether Innocent simply has in mind the first paragraph (Te igitur) of the present Roman canon (which also ‘commends’ the offerings) there can be no doubt that c. a.d. 400 the ‘Naming’ of the offerers at Rome comes in approximately the same place as at Milan, in the eucharistic prayer itself.

As now arranged the canon runs as follows: After the whispered offertory prayer by the celebrant (and the preface and sanctus, which in their present form are a later interpolation not contemplated by Innocent I), (Cf p.539), the prayer opens abruptly:

‘We therefore humbly pray Thee, most merciful Father, through Jesus Christ Thy Son our Lord and beseech Thee that Thou wouldest hold accepted and bless these gifts (dona), these ‘liturgies’ (munera), these holy and unspotted sacrifices: which first we offer for Thy holy catholic church, that throughout all the world Thou wouldst be pleased to give her peace, safety, unity and Thy governance:

‘Together with Thy servant N. our Pope [and N. our bishop and all the orthodox and the worshippers (who are) of catholic and apostolic faith] remember Thy servants and handmaids N. and N. and all who stand around, whose faith is accepted of Thee and whose devotion known [for whom we offer unto Thee, or] who offer unto Thee this sacrifice of praise, for themselves and all who are theirs...’

Just as the Mozarabic rite with its ad nomina offertory prayer still preserves the ‘naming’ of the offerers at the same point of the rite as in the days of the Council of Elvira c. a.d. 305, viz. at the offertory; so the Roman rite equally seems to preserve the position of the ‘naming5 customary in Italy c. a.d. 395, viz. soon after the offertory, in an early passage of the eucharistic prayer itself. Which of these represents the older tradition in the West is a point on which opinions will probably differ.

We may note here two points: 1. That whereas in Sarapion and at Jerusalem and probably in Africa, the only names read out appear to be those of the dead; at Rome and in Spain, so far as the evidence goes, the only names anciently read out were those of the living. And in fact it has been demonstrated that the commemoration of the dead which now appears as an invariable paragraph of the Roman canon, though it is genuinely ancient and of Roman composition, was originally only inserted in that prayer at funerals (and requiems generally), and formed no part of the Roman rite on other occasions. Its transformation from an occasional to an invariable part of the canon began in France in the eighth-ninth century, and was not accepted at Rome until the ninth-tenth, and in some Italian churches not till the eleventh century. 2. It is also plain that all this early evidence contemplates only the reading out of names of strictly local interest, whether they are those of living communicants or of deceased members of the local or neighbouring churches. The ‘Names’ are in fact the ‘parochial intercessions.’

In all this, however, we have met nothing which quite corresponds to the Eastern ‘Diptychs.’ These were two conjoined tablets, the one containing the names of living persons to be prayed for, the other containing a list of saints commemorated and of the dead persons recommended officially to the prayers of the church. It is first and foremost this combination of lists of the living and dead which distinguishes the ‘diptychs’ proper from the various customs of ‘naming’ which we have just been studying.

The diptychs come into sudden prominence at Constantinople c. a.d. 420 in the course of the disputes which took place there over the insertion or omission of the name of S. John Chrysostom, the ‘deposed’ bishop of that city who had died in exile in 407. From the official correspondence with other churches which arose about this we learn that at that time at Constantinople the diptychs (1) comprised separate lists of names, of the living and dead; (2) that each list was arranged in ‘ecclesiastical’ precedence, bishops first, then other clergy and finally laity; (3) that the whole succession-list of past bishops of Constantinople was included in the diptych of the dead, while the list of dead emperors headed the departed laity. It is clear also that at Antioch and Alexandria there were then diptychs of some kind, or at least lists of the dead. From the fact that these two churches were urged (and in the one case agreed and in the other indignantly refused) to follow Chrysostom’s own church of Constantinople in inserting his name among the dead, it is clear that some non-local names (besides departed emperors) must have been included in the case of the two southern churches; one would expect it to have been the same at Constantinople, though the evidence does not actually make this clear. But of the principle upon which such foreign names were selected — and some selection was necessary if the lists were not to grow intolerably long — we learn nothing.

From this time onwards, and especially down to c. a.d. 600, the diptychs are constantly in question in the East in connection with ecclesiastical politics, and accusations and counter-accusations of heresy. From the fifth century onwards the four great Eastern sees were supposed each to name the reigning patriarchs of the others in their diptych of the living. But in the interminable disputes and alliances and counter-alliances of patriarchates which went on under theological pretexts in this period (in all of which the question of the centralization of the political control of the East at Byzantium was seldom far from anyone’s mind), the solemn insertion or erasure of names and sees in the diptych of the ‘living was little more than a public register of how the political position stood at the moment. The confusion was just as great in the diptych of the dead. As the political balance between Melchites (‘King’s Men,’ as the orthodox were called) and Monophysites (or federalists) swayed, royalists and heretics succeeded one another in the same bishopric, and solemnly re-inserted, or ejected with anathemas the names of their predecessors in the local diptychs. The name of Dioscorus, the monophysite patriarch of Alexandria condemned by ‘the emperor’s Council’ (as both heretics and orthodox termed the Council of Chalcedon in the East) was removed from the Alexandrian diptychs by his orthodox successor Proterius, the nominee of the Byzantine government. When Proterius in turn was murdered by his monophysite successor, Timothy ‘the Weasel,’ the name of Dioscorus was restored, and that of Proterius removed with execration at the very moment when Constantinople was loudly numbering him among the martyrs. Names were removed or reinserted wholesale in some churches, according as the dead bishops had or had not agreed with the living one. Bishop Peter of Apameia in Syria removed the names of all his predecessors for some fifty years back at one stroke. Nothing much less like an ‘intercession list’ than the diptychs in actual Eastern practice can be imagined.

Yet it seems certain that it was in this that they had originated. In the liturgy of S. James the diptychs have always stood at just that point of the rite at which Cyril of Jerusalem (a.d. 347) mentions the ‘naming’ (of the dead only) — in the intercessions after the consecration; and they stood at the same point in the Antiochene and Constantinopolitan rites c. a.d. 390-400. But it is noteworthy that as at Jerusalem in Cyril’s time, so at Antioch and Constantinople fifty years later, only the dead are spoken of as being actually ‘named’ individually; and those named are very clearly, from what both Cyril and Chrysostom say, the ‘parochial’ dead, those personally known and loved and mourned by members of the congregation, together with a list of the past bishops of the see. But it is entirely clear from the course of the disputes about the insertion of the name of Chrysostom c. a.d. 420 that the lists by then had assumed a somewhat different character. They were now officially compiled, and admittance to them implied something more than just being dead. It was a judgement of the orthodoxy and of the eminence of the departed. It would appear, therefore, that the diptychs, in the form they finally took in the East — i.e. a combination of lists of both living and dead persons — and for the purpose they came to serve in the Greek churches — i.e. an index of ecclesiastical politics — are a development of the church of Byzantium in the years between c. a.d. 405 and 420. When one considers the peculiar state of that particular church at that time, as it is described in the lively but disillusioning pages of the Byzantine layman Socrates, I for one am not entirely surprised.

Whether in the properly Greek churches amid all this clash of great names and high policy the ordinary parochial dead — the communicants or the presbyters and deacons who did the pastoral work that must have gone on — ever got remembered in the diptych of the dead by name, we have not sufficient evidence to decide. So far as the great churches are concerned it is very unlikely; in the countrysides it may have been different. Nor do I see anything to suggest that the names of the living communicants (as in the West) or subjects for parochial intercession like the names of the sick, were ever entered on the Greek diptych of the living. That was reserved for the emperor and his family, the patriarchs of the great sees and the local bishop. First and last, the Greek diptychs properly so called have always been what they already are when we first hear of them at Constantinople c. a.d. 420 — instruments of strife in high places and not much more.

The entry of the four oecumenical councils, inserted (rather oddly) into the diptych of the dead by the church of Constantinople in a.d. 518, removed the diptychs further than ever from the notion of an intercession list. But this too was a political move. The Byzantine (or ‘centralising’) party had just recovered control of this see from the Monophysites who rejected the Council of Chalcedon. It was at first proposed to insert in the diptych only a commemoration of the first two general councils, which everyone accepted. The Byzantines changed their minds and inserted the fourth (Chalcedon) and with it the third, not so much because it was orthodox (they had themselves originally proposed the commemoration of the first two councils only) as because they realised that this would affirm the renewed opposition of the church of Constantinople to the federal claims of Syria and Egypt, which rejected the fourth Council.

The case is rather different with the history of the diptychs in the Egyptian and East Syrian rites.

In Egypt in the liturgy of S. Mark there is in fact only one diptych in which individuals are named — that of the dead. It consists, as in Sarapion, simply of a list of names read out in the course of the eucharistic prayer, which cannot be set down in the liturgical MSS. because it is not an ‘official’ list at all, as in the Greek diptych. It varied from church to church and from month to month, the names entered being those of the ‘parochial’ dead. But though the Alexandrian rite has thus retained exactly the form of the ‘naming’ of the dead found in Sarapion’s rite c. a.d. 340, it has shifted its place, and appended it to the lengthy intercessions ‘by categories’ for the church and the world — for the living — which it places after the first paragraph of the eucharistic prayer. This is precisely the point at which the Roman and Italian evidence of the late fourth and fifth centuries places its (much less developed) intercession for the church followed by the ‘naming’ of the living. (This is one of several coincidences of structure between the Alexandrian and Roman rites c. a.d. 400 which deserve more attention than they have received in modern study.)

The coincidence may well have been even more striking in some Egyptian country churches than at Alexandria itself. The chance discovery of a seventh century Egyptian diptych from the region of Thebes in Upper Egypt reveals that there, besides the patriarch of Alexandria and the local bishop, it was precisely the living communicants, the ‘offerers,’ who were named, as in the West: ‘And for the salvation of the most pure clergy standing around and the Christ-loving laity; and for the salvation and bodily health of the offerers so-and-so (masc.) and so-and-so (fem.) who have offered their oblations this day, and of all offerers’ (i.e. of all who are regular communicants, but are not ‘offering’ at this particular celebration).

In the East Syrian rite what Brightman calls ‘The Diptychs’ are read out at the offertory — as in the Mozarabic rite. There are two ‘books,’ of the living and dead, not quite the same in character. That of the living is brief and general in its contents, a summary of categories of people, in which the only individuals mentioned by name are the Nestorian patriarch and the local Nestorian bishop. The ‘book of the dead,’ on the other hand (with various alternative forms) takes up approximately eleven times as much space as its fellow in Brightman’s print. It consists of long lists of proper names, which include not only the great saints of the Old and New Testament and the succession-list of the Nestorian patriarchs of Mesopotamia, but all sorts of local worthies like ‘Rabban Sabha and the sons of Shemuni who are laid in this blessed village’ and ‘the illustrious among athletes and providers of churches and monasteries, generous in alms, guardians of orphans and widows, the Emir Matthew and the Emir Hassan and Emir Nijmaldin who departed in this village.’ And these loving and intimate local details vary from one MS. to another in a way that the stereotyped form of the Greek diptychs has never varied. It is clear, I think, that while the East Syrian diptych of the dead represents a genuine survival of the Naming’ of the ‘parochial’ dead, known and mourned by the congregation, the diptych of the living on the contrary represents an imitation of the formal Greek practice, inserted in the period when it had come to be taken for granted that there ought to be two diptychs.

It is clear from Narsai that in the later fifth century the East Syrian rite already contained both diptychs in much their present form. But the little prayer which according to him the people add after them runs thus: ‘On behalf of all the catholici (Nestorian patriarchs), on behalf of all orders deceased from holy church; and for those who are deemed worthy of the reception of this oblation, on behalf of these and of Thy servants in every place, receive, Lord, this oblation.’ In most rites the people’s prayers have a way of being more archaic than the clerical formulae they accompany. This prayer would suggest that it originally followed a ‘naming’ of the dead (headed by a succession-list of the Nestorian patriarchs) which was not preceded by a diptych of the living; and that if any living persons were subsequently ‘named,’ they were the communicants, as in the West.

It may be said, What then of the Western diptychs? What of those famous Roman diptychs, which as a number of modern scholars (beginning with Bunsen a century ago) have pointed out, resulted in the ‘dislocation’ of the Roman canon? What of the old Irish diptychs in the Stowe Missal, which once proved (somewhat inadequately) the ‘non-Roman’ character of the Celtic rites? What of the Mozarabic rite, which it has been said ‘has retained the diptychs in the position they originally occupied in all the primitive rites’? What of actual surviving ivory tablets which served this purpose, like that containing the list of the early bishops of Novara now at Bologna, or the Barberini diptych in the Louvre? In order to ascribe one of the institutions most venerated by generations of liturgists to a comparatively late initiative of the Byzantine church, are we not overlooking the multitudinous evidence that the West also, Rome included, once had diptychs?

I think we must distinguish carefully what we mean by ‘diptychs.’ If we mean simply lists of ‘names’ read out at the eucharist, whether of the communicants or (alternatively) of the dead, then the West had these customs before ever Constantine came to Constantinople. But if we mean that combination of lists of the eminent living and dead, officially drawn up and regulated from time to time by the higher ecclesiastical authorities, which is what ‘the diptychs’ were understood to mean by the church of Constantinople when it first instituted them, then the West never had any ‘diptychs’ properly so called at all. In the fifth and sixth centuries there was a tendency to copy some of the new Eastern fashions in this matter in many Western churches, including Rome. But it will be found upon examination that it was Syrian rather than Byzantine customs which chiefly proved attractive.

The Roman ‘diptychs’ are a myth. The most prominent feature of the Byzantine diptych of the living was the commemoration by each great church of the reigning patriarchs of all the other patriarchal churches. The local Roman ‘naming’ of the living at the beginning of the canon never mentioned any prelate whatever except the local bishop, the Pope. In the occasional pothers about the insertion of the Pope’s name in the Eastern diptychs, when communion was restored after a schism at various times from the fifth to the eighth centuries, there was never a suggestion by either side that Rome should return the compliment. Both parties knew that the Roman rite contained no opportunity of doing so, having retained the old purely local or parochial character of its ‘naming’ of the living.

As for the diptych of the dead, it did not exist at Rome. Edmund Bishop has shewn by the irrefutable evidence of the earliest extant MSS. of the Roman canon that the commemoration of the dead now found in the second half of that prayer has had a somewhat involved history. It is ancient and of genuinely Roman composition, but at Rome itself down to the ninth century it formed no part of the Roman canon as recited in the public masses of Sundays and festivals. It was a peculiarity of funeral and requiem masses, like a ‘proper’ preface for a festival, only inserted on specifically funerary occasions. Its use as an invariable part of the canon on all occasions begins in Frankish Gaul in the seventh-eighth century. The particular recension of the text which was eventually adopted betrays the hand of those Irish monks who in so many matters are at the bottom of Western innovations in liturgy during the dark ages. The Roman church only began to adopt this French novelty of commemorating the dead in all masses during the ninth century, and Italian MSS. of the Roman rite which did not allow for the new fashion were still being copied in the eleventh century. The absence of any ‘naming’ of the dead whatever in the authentic Roman rite on ordinary occasions is one contrast with the Eastern diptychs, which as we have noted owe their origin to the ‘naming’ of the dead in the fourth century Jerusalem liturgy. The purely local and parochial character of the ‘naming’ of the living in the Roman rite, by contrast with the international and diplomatic emphasis of the ‘naming’ of the living in the Byzantine diptychs, is another. Between them they make any application of the Byzantine term ‘diptych’ to the Roman ‘commemorations’ wholly misleading.

In so far as the alleged ‘dislocation’ of the Roman canon does not arise from mere modern misunderstandings of the tenor of its exceptionally archaic prayers, its cause must be looked for chiefly in things like the clumsy insertion of the sanctus in the fifth century. But there is one element connected with the origins of the diptychs at Jerusalem, which has had some effect. Cyril’s account of the ‘naming’ of the dead there c. a.d. 348 mentions the saints and a catalogue of the dead bishops of Jerusalem, as well as the more ordinary ‘faithful departed.’

We have seen that a ‘naming’ of martyrs in the eucharistic prayer had been adopted in S. Augustine’s rite (from Jerusalem) c. a.d. 410, though we hear nothing of a catalogue of the dead bishops of Hippo. The ‘naming’ of the saints in the eucharistic prayer was adopted at Rome, somewhat awkwardly and in a rudimentary fashion, apparently in the time of that innovating pontiff, Pope Gelasius (a.d. 492-6). The lists were elaborated by haphazard additions during the sixth century, but their present arrangement appears to date only from the reforms of Pope S. Gregory I, c. a.d. 595. By contrast with the deliberately ‘international’ character of the lists of saints in the later Jerusalem diptychs, the Roman lists never quite lost their old-fashioned parochial character. The list of men martyrs still includes only four non-Roman names out of sixteen; and one of these, Ignatius of Antioch, had been martyred at Rome. The women martyrs contain four foreigners out of seven, but two of them, the Sicilians Agatha and Lucy, were introduced (almost certainly) by S. Gregory himself in his final revision; and the Balkan Anastasia, though her popularity was chiefly due to Greek settlers at Rome, seems to have got into the canon by confusion with the ‘Anastasia’ who had built an old Roman parish church, the titulus Anastasiae. The Roman canon never adopted the other Jerusalem innovation of a catalogue of the past bishops of Rome, despite the occurrence of the names ‘Linus, Cletus, Clement, Sixtus,’ which has been supposed to be the relics of one. It appears probable that Pope Sixtus II (martyred a.d. 258) had been commemorated in the canon for about a century before the name of Clement was added, and that Linus and Cletus were only inserted later still by S. Gregory I, in the final revision.

In the Irish Stowe Missal, however, there is a ‘diptych of the dead’ (though not one of the living) fitted into the text of the Roman canon. It contains a long list of Irish names, the owners of which with one doubtful exception — Maelruen — had all died before a.d. 739. (It may be remarked that the diptych had thus received at the most one addition in the century before the present MS. was copied, which suggests that the diptych of the dead was not a very living institution in the Irish rite.) But a comparison of the Stowe diptych with that found in the Mozarabic rite will, I think, convince anyone of the origin of this supposed Irish practice. The Spanish diptych has been ‘localized’ and adapted in the usual Irish way; but the Irish document is not a native product, but a direct copying of Spanish custom.

And if one wishes to pursue this ‘key’ Western diptych of the dead in the Mozarabic rite to its source, a comparison of it with the diptychs of the dead in the Syrian rites will at once supply the solution. The Mozarabic document is simply an adaptation of the Syrian custom. Some Western churches adopted the Jerusalem custom of reciting a complete succession-list of their bishops in the eucharistic prayer, which in some cases lasted as late as the fifteenth century. What does not exist is any Western example of the specifically Byzantine custom of naming the chief foreign bishops with whom the local bishop was in communion in the diptych of the living.

It may be well to sum up here what is known about the practice concerning public intercessions in the course of the liturgy in various churches, since nothing has caused more confusion in the various manuals of liturgical history, and far-reaching (and quite erroneous) theories have been based upon those confusions.

We have to distinguish clearly as regards origins between the intercessions at the synaxis and those at the eucharist. Those at the synaxis were offered for categories of persons, the only individuals mentioned by name being the local Christian bishop and the Roman emperor. They took the form of a bidding, a pause for silent prayer and a collect. In the fourth century at Jerusalem these ‘prayers of the faithful’ were transferred bodily from the end of the synaxis to after the consecration, and made a part of the celebrant’s eucharistic prayer. This innovation was afterwards widely imitated in the East. In Christendom as a whole these prayers at the end of the synaxis suffered an eclipse during the fifth century. In the West they disappeared altogether (except for special survivals in Holy Week). In the Byzantine church they were replaced by the new Antiochene fashion of litanies. Later imitations of this Byzantine novelty in some sort re-introduced traces of them into some Western rites (e.g. the Mozarabic) in the form of litanies. But a comparison proves in every case, I think, the dependence of these later Western litanies on the Constantinopolitan text, and forbids us to treat them as any sort of authentic survival of the ancient Western pre-Nicene ‘prayers of the faithful.’ These had been dropped from the Western rites perhaps a century before the first Western imitations of the Byzantine litany made their appearance. Only in the Roman rite on Good Friday, and in the Egyptian liturgy of S. Mark, do the pre-Nicene ‘prayers of the faithful’ still survive in something like their original form.

The pre-Nicene intercessions at the eucharist proper were much more personal than those at the synaxis, and ‘named’ specific individuals. In Spain and Italy these were names of living persons, chiefly the communicants, and the same may have been the case originally in Syria and in some Egyptian churches. But in the oldest known Egyptian rite (Sarapion), and at Jerusalem, Antioch and Constantinople, the first evidence we have is of the ‘naming’ of the dead only, and the same appears to be true of pre-Nicene Africa. There is good ground for thinking that the original position of these lists of ‘names’ of the living was at the offertory, at all events in the West. The ‘naming’ of the dead after the consecration at Jerusalem may conceivably have been transferred to that point from the offertory at the same time as the intercessions from the synaxis, but there is no evidence on this. And in Africa in pre-Nicene times the dead were ‘named’ in the course of the eucharistic prayer.

The first combination of lists of names of living and dead, ‘the diptychs’ properly so-called, was made at Constantinople in ±e early fifth century, but these did not so much replace the old intercessions (now attached to the eucharistic prayer) as fulfill new official and diplomatic purpose. Outside the properly Greek churches, in East Syria, Egypt and the West, the older custom of ‘the Names’ continued in force at the old position, at the offertory. Imitation of Byzantium brought about the partial adoption of the form of diptychs in Syria, whence it spread to some Western churches in the sixth-seventh century. But in the non-Greek churches these imitations of the Greek diptychs always retained a ‘parochial’ and local interest, by contrast with the purely official character of the Byzantine custom. In the non-Greek rites, after the fusion of synaxis and eucharist these very ancient lists of ‘Names’ coming after the offertory in some sort supplied for the loss of the old ‘prayers of the faithful’ before the offertory, though they have no original connection with them. The ‘prayers of the faithful’ were the intercessions of the synaxis, the ‘Names’ were the intercessions of the eucharist, in the days when these were still two separate rites.

Just as the period of the ‘second stratum’ equipped the rite of the synaxis with a wholly new Introduction, so it equipped the rite of the eucharits with a wholly new Conclusion. Just as the tendency of the synaxis was to prefix new items before the old nucleus, so the tendency at the eucharist was to append them, leaving the old core of the rite relatively unchanged in both cases. The ‘second half of the eucharistic prayer had begun to be added to the original ‘thanksgiving series’ in the second century, and various additional items and paragraphs had been appended and then fused into the prayer in the course of the third and early fourth century. Then at Jerusalem, by the time of Cyril, the Lord’s prayer had been appended to the whole. Its independent existence as a prayer outside the eucharist had secured for this last addition that it should be allowed to remain as a separate item, and not be fused into the eucharistic prayer itself, as had happened to previous additions to that prayer. But even in this case we have seen that at Milan the Lord’s prayer was for a while placed between the body of the eucharistic prayer and its concluding doxology.

In the fourth century the tradition that in the rite of the eucharist proper there could be only a single prayer — ‘the’ prayer, the eucharistia — was beginning to break down. Supplements come to be made in this period which are no longer incorporated perforce into ‘the’ prayer itself, but are separate items. One such is the separate offertory prayer (of the ?5th century). This puts into words the meaning of the offertory, which the pre-Nicene church had been content to express by the bare action. Other such separate prayers were added even earlier to put into words the meaning of the fraction and the communion, which formerly the church had also been content to leave to speak for themselves.

There appears, too, for the first time something which would one day become the keynote of mediaeval and modern eucharistic devotion, the idea of special prayers in preparation for the individual act of receiving communion. However strange it may seem to us, this is an innovation in the fourth century. The old rite of offertory, prayer, fraction and communion had been unable to express this ‘communion devotion,’ except in the course of the eucharistic prayer. We can see the beginnings of this in Hippolytus (k), but here the emphasis is still on the corporate effects of communion — that all the communicants ‘may be made one.’ Sarapion (e2) strikes a new note: ‘Make all who partake to receive a medicine of life ... not for condemnation, O God of truth .. .’ It remains to be seea how this is amplified outside the eucharistic prayer itself.

Sarapion has no trace of the Lord’s prayer after the eucharistic prayer, but continues at once from its closing doxology and Amen with the rubric:

‘After "the" prayer the fraction and in the fraction a prayer.

‘Account us worthy even of this communion, O God of truth, and make our bodies to compass purity and our souls prudence and knowledge. And make us wise, O God of compassions, by the partaking of the Body and the Blood, for unto Thee through Thy Only-begotten (is) glory and might in holy Spirit. . .

After giving the fraction to the clerics, laying on of hands [i.e. a blessing] on the people:

I stretch out the hand upon this people and pray that the hand of the truth may be outstretched and a blessing be given unto this people through Thy love of men, O God of compassions, and through the present mysteries. May a hand of piety and power and discipline and cleanness and holiness bless this people, and continually preserve it to advancement and progress: through ...

‘After the distribution to the people a prayer.

‘We thank Thee, Master, that Thou hast called those who have strayed, and hast taken to Thyself those who have sinned, and set aside the threat that was against us, granting mercy by Thy loving-kindness and wiping (that threat) away by repentance and casting it off by the knowledge (that leads) to Thee. We give thanks to Thee that Thou hast granted us communion of (the) Body and Blood. Bless us, bless this people, make us to have a share in the Body and Blood: through Thy Only-begotten Son, through Whom . . .’

[There follows a blessing of oil and water offered for the sick, followed by the final blessing of the eucharistic rite.]

‘Laying on of hands after the blessing of the water and oil:

‘0 God of truth that lovest mankind, let the communion of the Body and the Blood go forth along with (symparabaineto) this people. Let their bodies be living bodies and their souls be clean souls. Grant this blessing to be a safeguard of their communion and a security to the eucharist that has been held. And beatify them all together and make them elect: through Thy Only-begotten Jesus Christ in ‘holy Spirit,’ both now and for ever and world without end. Amen.’

The old eschatological note is almost entirely missing from all this, only appearing in the last sentence of the final blessing. For the rest it is recognisably the ‘modern’ feeling of sacramental devotion that it expresses, concentrated on reception. The prayer at the fraction (carefully distinguished from ‘the’ prayer) shews that the fraction is still looked upon as a mere utilitarian preparation for communion, not a dramatic or symbolical act. Particular attention may be called to the blessing of the people before communion, which is found in all rites by the end of the fourth century. Its pointed bestowal upon the people after the clergy have made their communion suggests that it is a symptom of that increasing feeling that the ‘profane’ laity ought not to communicate, which soon led to their general abstention. It is designed to encourage and fit them to receive. The discerning reader who compares the things asked for the communicants in Hippolytus k (p. 151) or Addai and Mari i(p. 179) with those in Sarapion (e2 p. 164) and the prayers on p. 512, may detect the beginnings of a new psychological attitude towards the act of communion in the fourth century.

Cyril of Jerusalem has a different system, though its emphasis is the same. He does not mention the fraction and there is no blessing of the people, so that Sarapion’s first two prayers have no equivalent in the contemporary Jerusalem rite. Instead, the Lord’s prayer, with its petitions for daily bread and the forgiveness of trespasses ‘as we forgive’ (both of which Cyril explicitly interprets as a preparation for communion), acts as the people’s preparation. Then comes the bishop’s invitation ‘Holy things for the holy’ and the people’s reply ‘One only is holy.’ The people receive communion with bowed heads ‘as adoring and worshipping,’ and answer ‘Amen’ to the words of administration. Meanwhile a solo singer chants Psalm xxxiv, with its refrain or chorus ‘O taste and see how gracious the Lord is.’ Finally they are bidden, ‘While you wait for the prayer give thanks to God Who has accounted you worthy of such great mysteries.’ Evidently there was now a post-communion prayer at Jerusalem, but whether it corresponded more closely to a ‘thanksgiving’ or a ‘blessing’ is doubtful in view of the other Syrian evidence.

The North Syrian ‘communion developments,’ as represented by Ap. Const., viii. (supported in some points by the Antiochene writings of Chrysostom) are different again. There is no Lord’s prayer, but immediately after the doxology which concludes the intercessions attached to the eucharistic prayer the bishop greets the church ‘The peace of God be with you all? to which the people answer ‘And with thy spirit.’ There follows a series of ‘proclamations’ by the deacon, ‘bidding’ the people to pray for various objects, some connected with their communion, others being brief intercessions. There is no direction that the people shall answer Kyrie eleison, but the whole has the appearance of a litany. It was apparently during this bidding that the fraction took place. There follows a solemn blessing of the people by the bishop: CO God the mighty and mighty-named, mighty in counsel and powerful in deeds, God and Father of Thy Holy Servant (pais) Jesus our Saviour: Look upon us and upon this Thy flock which Thou hast chosen through Him unto the glory of Thy Name. And hallow our bodies and soul (sic); make us worthy, being purified from all defilement of flesh and spirit, to receive of these good things here lying before Thee; judge none of us unworthy, but be Thou our helper, our succour and defender, through Thy Christ with Whom unto Thee be glory, honour, praise and laud and thanksgiving with the Holy Ghost for ever. Amen.’ (Even more clearly than in Sarapion this blessing is an encouragement and preparation of the communicants.)

The deacon then cries ‘Let us attend’ and the bishop gives the invitation — ‘Holy things unto the holy’ to which the people reply with a sort of prose hymn, There is one holy, one Lord Jesus Christ to the glory of God the Father, blessed for ever. Amen. Glory be to God on high, and in earth peace, goodwill towards men. Hosanna to the Son of David. Blessed is He that cometh in the Name of the Lord. God is the Lord Who hath shewed us light. Hosanna in the highest.’

The communicants answer ‘Amen’ to the words of administration — The Body of Christ,’ and The Blood of Christ, the cup of life.’ Meanwhile Psalm xxxiv is chanted, as at Jerusalem. There follow (i) a thanksgiving prayer (which rather wanders from the point into a repetition of the intercessions) and (2) a lengthy blessing, after which the deacon’s ‘Depart in peace’ dismisses the people.

The Antiochene rite as described by Chrysostom does not altogether support Ap. Const, in its details. The Lord’s prayer is said at Antioch as at Jerusalem. The psalm sung during the communion is cxlv (in the English numbering, cxliv in that adopted by the primitive church) which is certainly no less appropriate. After the communion there is a thanksgiving prayer, but it appears to have been of recent introduction, since Chrysostom has some difficulty in persuading the people to remain for it. He compares those who hurry out at the communion (the ancient completion of the rite) to Judas bursting out of the upper room on his mission of betrayal, and those who remain to his fellow-disciples awaiting the psalm at the end of supper and going out with the Lord. We shall meet again this inclination of the laity, based on traditional practice, to treat the post-communion as an optional ‘extra.’ He has nothing about a final blessing after the thanksgiving. The conclusion is still the deacon’s ‘Depart in peace.’

Theodore’s rite at Mopsuestia, however, omits the Lord’s prayer like Ap. Const.y viii. Evidently this Jerusalem innovation had not yet reached his church. The fraction follows immediately upon the intercessions that conclude the eucharist prayer and is done ‘so that all of us who are present may receive (communion).’ This is accompanied, as in Sarapion, by a prayer of ‘thanksgiving for these great gifts’ (Ibid., p. 105), and a blessing of the people. Then comes the ‘signing of the Body with the Blood,’ and the placing of a portion of the Host in the chalice. Then the deacon says ‘We ought to pray for those who presented this holy offering.’ ‘The priest finishes the prayer by praying that this sacrifice may be acceptable to God and that the grace of the Holy Spirit may come upon all, so that we may be able to be worthy of its communion, and not receive it to punishment,’ and again blesses the people (Ibid., p. 108). Then follows the invitation ‘Holy things ...’ and its answer. Then follows the communion received with ‘adoration’ and ‘fear’ — the actual phrasing of Cyril’s Catecheses obviously inspires Theodore’s instructions at this point — but there is no mention of an accompanying psalm. ‘After you have received... you rightly and spontaneously offer praise and thanksgiving to God.... And you remain, so that you may also offer thanksgiving and praise with all, according to the rules of the church, because it is fitting for all those who received this spiritual food to offer thanksgiving to God publicly...’ (Ibid., p. 114). There is nothing about a final blessing or the deacon’s dismissal in Theodore.

It is clearly the North Syrian rite of Chrysostom’s time which has governed the ‘communion devotions’ and post-communion of the present Byzantine rite, but the exact form does not appear in any of our extant North Syrian sources. There is the double blessing (as in Theodore) with the Lord’s prayer between them (as in Chrysostom). The deacon’s proclamation ‘Let us entreat on behalf of the holy gifts that have been offered and hallowed’ is followed by other intercessions, and the priest’s prayer for worthiness of reception ‘and not Unto judgement’ and the gift of the Holy Ghost, are as in Theodore.

The manual acts (fraction, etc.) have been complicated by the peculiar and mysterious addition of the pouring of a few drops of hot water into the chalice by the deacon (known as the zeon or ‘living water’). This ceremony is found only in the Byzantine rite, but it appears to be ancient there. In the sixth century the refractory Armenian patriarch Moses when summoned to Constantinople to appear before the emperor Maurice is reported to have answered, ‘God forbid that I should cross the River Azat or eat leavened bread or drink hot water.’ Since the second of his disinclinations reflects on the Byzantine eucharist, the third may very well refer to the zeon as an already established Byzantine peculiarity. The Greek devotional tradition explains it variously as symbolizing ‘the fervour of faith’ or ‘the descent of the Holy Ghost.’ But these are explanations devised for an existing traditional practice, not its originating cause — as to which, however, I am unable to make any suggestion.

The communion in the Byzantine rite is now accompanied by two chants, the one ‘Blessed is He that cometh in the Name of the Lord, God is the Lord who hath shewed us light’ (cf. Ap. Const., viii.), the other a Byzantine ‘prose hymn’ with a peculiarly striking melody:

O Son of God, take me this day for a partaker

Of Thy mystic supper, For I will not tell Thy secret to Thine enemies,

I will not betray Thee with a kiss like Judas

But like the thief confess Thee;

Immediately after the communion there is a further blessing of the people with the consecrated sacrament, of which a good deal is made in the Byzantine liturgical commentaries, in which it is said to symbolize our Lord’s blessing of His disciples at the Ascension. It is in fact a sort of substitute for communion devised to satisfy Byzantine non-communicant eucharistic piety.

The choir then sings the ‘departure chant’ of the day, a variable chant corresponding to the Western communion chant. The ‘thanksgiving’ proper has disappeared from the modern rite, though a short thanksgiving prayer was still found here in the ninth century, in the liturgies both of S. Basil and S. Chrysostom. All that is now left is a truncated version of a diaconal litany which formerly preceded the thanksgiving, followed by the dismissal ‘Let us depart in peace,’ said by the deacon. The ‘prayer behind the pulpit’ (opisthambonos), for which the priest comes out of the sanctuary, represents a sort of ‘conducted devotion’ after the service rather than an integral part of the rite, though its opening sentence fulfils also the purpose of a departure-blessing.

The Eastern communion devotions and thanksgiving are thus a product of the fourth century, and their development may be said to have been completed in principle by a.d. 400. In the West, development was less rapid.

We have seen that the singing of a psalm during the communion was a novelty in Africa (Cf. p. 492; adopted from Jerusalem?) early in the fifth century, and does not appear to have been taken up elsewhere in the West for some time after that. A letter of Augustine’s written c. a.d. 410 rather before the adoption of this novelty gives us the order of the prayers of the eucharist ‘which’ he says ‘every church or almost every church customarily observes.’ There is a prayer ‘before that which is upon the Lord’s table begins to be blessed’; (the ‘prayer of the day,’ or an ‘offertory’ prayer?). There follows the eucharistic prayer ‘when it is blessed and hallowed and broken for distribution, which whole prayer almost every church ends with the Lord’s prayer.’ (It is interesting to find the prayer at the fraction and Lord’s prayer included within the eucharistic prayer; it shows how the appending of items to ‘the’ prayer was understood not to violate the old rule that the eucharistic prayer proper must be a single whole.) The kiss of peace followed at this point. Then ‘the people are blessed. For then the bishops like advocates present those whose cause they have undertaken before the most merciful judgement seat of God by the laying on of hands. When all this has been done and the great sacrament partaken, the thanksgiving ends all.’

It seems certain that there was less uniformity even among the Western churches at this time than Augustine supposes; but this outline probably holds good in the main for all the African churches at least. What is particularly interesting is to find the blessing of the people before communicating in the African rite at this early date. During the Pelagian controversy Augustine was accustomed to quote the custom of blessing the people, and also the contents of some of these blessings, as an argument against Pelagius. It seems to be assumed by those liturgical authors who have treated of the matter that it is always to this pre-communion blessing that Augustine is referring. I see no grounds for this assumption in the evidence, since this was not the only occasion when a blessing was given in public worship. It is unlikely — given his usual reserve about the contents of the eucharistic prayers — that Augustine would cite a eucharistic blessing, when others given at the more ‘public’ worship of the synaxis and the office were available to prove his point. But if the assumption is justified, it is important to note that Augustine cites more than one formula, adding on one occasion ‘and others like these’ (et caetera talia). If these are pre-communion blessings, then in Africa this had already become a variable formula, not a fixed one, as it remained in the East. (This would be one of the earliest suggestions we have of the introduction of variable prayers at the eucharist, which afterwards became a notable feature of all Western rites.)

At Rome also there appears to have been a blessing of the people before communion in the late fourth century. The mysterious Roman author who goes under the name of ‘Ambrosiaster’ (c. a.d. 385) tells us: The priests, whom we call bishops, have a form drawn up and handed down to them in solemn words, and they bless men by applying this to them... and though a man be holy, yet he bends his head to receive the blessing.’ This looks like a fixed form, as in the East, not a variable one as later in the West.

The custom had fallen out of the Roman rite by c. a.d. 500, but Dom Morin has suggested that some of the formulae have survived in the special oratio super populum (prayer ‘over the people’) now appended to the post-communion thanksgiving prayers in the Roman rite during Lent. These prayers (of which there is one for each week-day in Lent) are now not always distinguishable from ordinary collects in structure. But in a number of them the celebrant instead of praying with the people (in the first person plural) prays for them (in the third person plural), making a sort of ‘prayer-blessing’ (like the final blessing in Sarapion, p. 512), and this appears to be the original type. And they are preceded by a ‘proclamation’ by the deacon, ‘Bow down your heads unto the Lord,’ which is verbally identical with the deacon’s ‘proclamation’ before the pre-communion blessing in some of the Eastern rites.

Whether it be true that the Lenten oratio super populum at Rome is a survival (transferred to after the thanksgiving) of the blessing before communion, or whether this Lenten peculiarity has some other origin, the fact that the Roman rite in the fifth century always had a blessing before the communion appears to be certain. And it has this much importance, that it is one more little piece of evidence going to shew that in the fifth century all the Western rites formed a group and were similar in structure. For this pre-communion blessing perpetuated itself in some of the non-Roman Western rites, and persisted as a special local custom in many Western churches even after the adoption of the Roman rite.

Before discussing the Roman thanksgiving prayer it will make for clarity to turn to the early Spanish and French rites. The ninth century Spanish Liber Mozarabicus Sacramentorum presents us with the following preparation for communion. After the (variable) eucharistic prayer comes the fraction, then the creed, followed by the praefatio to the Lord’s prayer (varying in every mass), then the Lord’s prayer itself; after this, a threefold blessing and the communion. The following, for New Year’s Day, is likely to be one of the older compositions in the book and will serve for an example:

Praefatio: ‘O Lord Who art the great day of the angels and little in the day of men, the Word Who art God before all times, the Word made flesh in the fullness of time, created beneath the sun Who art the sun’s creator: Grant unto us the solemn assembly of the church’s dignity in Thy praise on this day (sic) that we who have consecrated the beginning of the year to Thee with these firstfruits, may by Thy grace sacrifice to Thee the whole time of its course by such ways and works as shall please Thee (totius temporis spatium tibi placitis excursibus atque operibus facias inmolari). For at Thy command we pray to Thee from earth, Our Father...’ There follows this threefold blessing, preceded by the deacon’s proclamation, ‘Bow yourselves for the blessing.’ ‘May all of you who welcome the beginning of this year with His praises be brought without sin to its ending by the abiding protection of our Saviour. Amen. And may the same our Redeemer so grant unto you that this year be peaceful and happy that your heart may ever be waiting upon Him. Amen. That blessed of Him Who made heaven and earth that which you now begin in tears you may afterwards fulfil with spiritual songs. Amen.’

What is interesting is to find that in the ninth century this threefold blessing is still the concluding text of the rite so far as the celebrant is concerned. Just as the Spanish rite at that date had developed no collect in the introduction before the old nucleus of the synaxis, so it had developed no thanksgiving-prayer after the old nucleus of the eucharist, but virtually ended with the communion. The fourth Council of Toledo in a.d. 633 (can. 18) had sternly reprimanded those who attempted to transfer the blessing from before the communion to after it, and had ordered that communion should end the rite as heretofore; and the Council’s legislation had evidently maintained the old ways for another 200 years in Spain. But when the Liber Sacramentorum was written the custom of saying a short public thanksgiving prayer (completuria) after the communion was just beginning to spread, doubtless through imitation of the Roman rite. Four days in the year are provided with a completuria in this MS.; more are found in those of the next century, and in the eleventh century Liber Ordinum, the majority of the masses are so equipped. But by that time the episcopal blessing before the communion was so unalterably fixed in Spanish tradition that it was never transferred to the end of the rite, even after the thanksgiving prayer had been added in the ninth-tenth century. It still remains where the fourth Council of Toledo fixed it, before the communion.

The developed Mozarabic post-communion of the middle ages runs thus: after the communion the choir sing an anthem (corresponding to the Roman communio) followed at once by a brief (variable) thanksgiving collect. Then the celebrant greets the church with The Lord be always with you’; R7. ‘And with thy spirit.’ The Spanish deacon’s ancient dismissal, ‘Mass is over,’ is amplified to ‘Our solemnities are completed in the Name of our Lord Jesus Christ. May our devotion be accepted in peace.’ R/. ‘Thanks be to God.’ This presents in its structure a close parallel with the early mediaeval Roman conclusion.

In Gaul the arrangement was slightly different. The fraction came immediately after the end of the eucharistic prayer; then followed the Lord’s prayer with its praefatio. Then the deacon sang ‘Bow down for the blessing’ and the bishop pronounced the threefold blessing divided by the people’s ‘Amens’ as in Spain. But in France the blessing in its full form became a special episcopal prerogative, and priests used a shorter and less elaborate form. The ‘episcopal benedictions’ became a special feature of the eucharist in France. Pope Zacharias already viewed them somewhat severely in a.d. 751 as ‘not according to apostolic tradition but done out of vainglory, bringing damnation on themselves, as it is written "If any one preach unto you another gospel than that which was first preached unto you, let him be anathema." Nevertheless, they survived the adoption of the Roman rite under Charlemagne, and lasted in many French churches into the eighteenth century, and at Autun into the twentieth.

A different set of these benedictions was provided for every liturgical day in the year, some of them with five, seven, eight or nine clauses, all divided by ‘Amens.’ From Gaul the custom spread to England, Germany, and even one or two Italian and Hungarian churches. In England they lasted down to the Reformation at the mass of a bishop or abbot.

The interesting thing is that the early French evidence reveals the same absence of a thanksgiving prayer as in the early Spanish rite. Not only does the Council of Orleans in a.d. 512 (can. 1) insist that the bishop’s blessing before the communion is the end of the rite, before which no one must presume to depart; but S. Caesarius of Aries in sermons preached about the same time reveals the difficulty of inducing the laity, now that they no longer communicated, to remain even as long as that. He repeatedly exhorts them to stay until ‘the whole mass’ (missas ad integrum) has been completed, which, he says, is not until the bishop’s blessing after the Lord’s prayer has been given. The earliest Gallican liturgical MS., the Masses of Mone, has added a short thanksgiving collect after the communion at all masses, varying with the day. But this is evidently a seventh century addition to the true Gallican rite, based on the Roman model. ‘Germanus,’ for what its evidence is worth, has still no prayer of any kind after the benediction before communion. But since it also prescribes the saying of the creed as a preparation for communion, which is a Spanish not a French custom, it is perhaps no longer representative of what had come to be the general contemporary French practice c. A.D. 700.

When the episcopal blessing before communion formed the final prayer of the rite, followed only by the communion (in which the majority of the laity no longer took part), it naturally came to be regarded as a sort of climax. What appears to have caused its removal at Rome is the introduction of a brief thanksgiving or post-communion prayer after the communion, parallel to the collect before the lections. When this was first adopted at Rome is obscure, but it must have been some time during the fifth century. Post-communions are provided as systematically as collects for all masses in the Gelasian Sacramentary, whose groundwork seems to date from c. a.d. 500. But I have been unable to discover any earlier reference to any thanksgiving prayer at Rome, and the analogy of the Spanish and French rites suggests the possibility that its adoption took place after that of the collect, perhaps only towards the end of the fifth century. Many of the postcom-munions themselves are hardly comparable with the collects in their workmanship, either for range of ideas or expression, which suggests that they may date from a rather different period, though both sets of prayers have the same structure.

The introduction of a concluding prayer after the communion may well have suggested to the Roman sense of form the idea of removing the solemn blessing, which had previously come before the communion, to after the thanksgiving (the position of the super populum in Lent). But the old tradition that the deacon’s dismissal ought to end the rite died hard at Rome. The blessing after the thanksgiving was in form hardly distinguishable from a second postcommunion collect and served no particular purpose, since it was still followed by the deacon’s Ite missa est. And so it was dropped altogether (except in Lent, a season when archaisms are apt to survive in all rites).

The brief blessing ‘May God Almighty, the Father, Son and Holy Ghost, bless you’ which now follows the Ite missa est in the Roman rite begins as a sort of informal piece of politeness. In the Ordo Romanus Primus, as the Pope goes out to the sacristy after mass other bishops step forward to ask his blessing, and he replies ‘May God bless us’ — a courtesy. The people bow to him as he passes through the congregation in procession to the sacristy, and he replies by signing them with the cross. Only in the eleventh century did priests as well as bishops begin to bless the people as they went from the altar. The custom spread, apparently from France, but not very quickly, during the middle ages. In England the mediaeval derived uses (e.g. Sarum, Westminster and Hereford) did not adopt it. The present forms (differing for priests and bishops) were not finally fixed in the Roman missal until the pontificate of Clement VIII in the seventeenth century.

The blessing is always said by priests in the Roman rite, even at a sung eucharist, an indication that it is no inherent part of the public rite, but rather to be classed with those semi-private devotions like the ‘last gospel’ (John 1:1-14) and the ‘preparation,’ which grew up as a sort of ‘third stratum’ during the middle ages, around the completed Shape of the Liturgy, rather than as part of it. The real end of the rite is communion and the deacon’s proclamation that it is complete. To this the ‘second stratum’ added a brief and formal thanksgiving — for how can public thanksgiving for such an intimate thing as the union of the soul with God be anything but formal? Even though it is right that we should all give thanks together for the same gift, it was also a right instinct which made it brief — a gesture only — and left the soul to its Lord, sending it back with Him to daily life with ‘Depart in peace’ or some such phrase. There is a certain ‘clericalism’ about reinforcing communion with a priestly blessing, however true it be that ‘the blessing of a good man availeth much.’ The primitive church rejoiced in such blessings and multiplied them, but she did not choose this particular moment for imparting them.

The Shape of the Liturgy as it stood c. a.d. 800 all over Christendom remains substantially intact henceforward, because it is an organically completed thing, logically adapted to express the eucharistic action it performs in a society which is nominally Christian in its assumptions about human life as a whole. It consists roughly of four parts: the introduction (added by the ‘second stratum’); the old nucleus of the synaxis (minus the concluding intercessory prayers which had atrophied in different ways in all rites); the old nucleus of the eucharist; and a brief appended thanksgiving (the other addition of the ‘second stratum’). Henceforward the additions and changes made have about them the character of mere decorations, rather than of structural changes, though they are numerous and various and continual enough in all rites.

A sufficient illustration is the history of the Agnus Dei in the Roman rite. This little prose hymn ‘O Lamb of God that takest away the sins of the world, have mercy upon us’ was first introduced into the Roman rite during the fraction immediately before the communion by Pope Sergius I (a.d. 687-701). This act of adoration of our Lord present in the sacrament (for it is intended as such) is a somewhat ‘un-Roman’ form of eucharistic devotion. But its origin is explained by the fact that Sergius was himself a Greek, born in Syria; the idea of ‘the Lamb of God’ had for centuries attracted a special devotion in the Syrian church. The Agnus Dei is in fact only one more of those ‘Syrian importations’ of which we have had other instances, which were constantly enriching Western forms from the fourth century to the eighth. As Sergius introduced it the hymn seems to have been sung twice, once by the choir, and then repeated by the people. This continued down to the eleventh century. In the twelfth century the liturgist Beleth describes it as being sung three times in French churches, with ‘grant us peace’ substituted for ‘have mercy upon us’ at the third repetition. At the end of the thirteenth century, however, it was still being sung at Rome without this new variation, a custom maintained to this day in the Pope’s cathedral, the Lateran basilica. Yet another little variation for use at funerals had made its appearance in France during the eleventh-twelfth century — the substitution of ‘grant them rest’ for ‘have mercy upon us.’ This in turn led to the further change of adding .’ . . rest everlasting’ at the third repetition at funerals, in those churches which substituted ‘grant us peace’ at the third repetition on ordinary days.

When a single small item of late introduction can go through so many variations in a few centuries it is obvious that the following of all the innumerable changes in the details of rites ceaselessly made and remade all through the middle ages would be a wearisome and lengthy business. Especially is this the case with the additions made by what I have called the ‘third stratum’ — c. a.d. 800-1100. These additions were not protected from change either by their structural usefulness, like those of the ‘second stratum,’ or by immemorial tradition like the primitive nucleus; and in consequence the persistent innovating tendency of the clergy in all ages with regard to details of the liturgy had comparatively free play with these, and also with minor matters of ceremonial whose development or alteration goes on persistently in all rites down to the sixteenth-seventeenth century. But these little matters of fashion and fancy change nothing in the main outline of the Shape of the Liturgy, which remains everywhere much as it was c. a.d. 800.

After this the persistent desire to improve upon the traditional liturgy, restrained by its ‘completeness’ as a whole, finds expression chiefly in the composition of ‘devotions’ preparatory to and looking back upon the liturgical action itself — ‘preparations’ and ‘thanksgivings’ — which are of the greatest interest from the point of view of the history of religious psychology, but are less closely connected with the history of the liturgy itself.

A few words must be said, however, of the instances of this tendency which have managed to attach themselves as official and prescribed parts of the rite in the liturgies of the East and West alike — the ‘preparation’ and various post-liturgical devotions.

In both the Byzantine and Roman rites there is now an officially prescribed ‘preparation,’ a series of devotions for the priest and his assistants, before the service begins. The forms now in use in both cases begin to take shape about the eleventh-twelfth century, and reach their present text in the sixteenth century. In both cases the process begins with the prefixing of a single private prayer to be said by the celebrant in the sacristy, which is found in both East and West as far back as the seventh century. Certainly private devotion had always exacted devout preparation from clergy and laity. But it is a different thing when official regulation begins to prescribe the form of private devotion, and to draw its exercise into the sphere of the fixed liturgical action instead of leaving it to its own natural field — the individual’s personal preferences under the action of grace. It presages a good deal that has subsequently taken place in the way of ‘psychologising’ the eucharist, and removing the emphasis from the corporate action to the individual’s subjective feelings and thoughts about that action.

It is further noticeable that whereas this earliest Eastern ‘prayer of preparation’ in the seventh century is concerned with the preparation of the elements, the earliest Western prayers (apologiae as they are called) are true to the inherent Western bent for ‘psychologising,’ and are wholly concerned with the preparation of the priest. This original difference of bias in the devotion of the two churches is something which has persisted, not only in their preparatory devotions but in their eucharistic devotion as a whole. Thus while the Eastern veneration for even the unconsecrated elements has made the great entrance one of the moments of supreme worship in the rite, the Western tendency has been to make of these psychological reactions of the individual not merely a preparation for the rite, but something which is of its very structure. We find these apologiae prefixed to the Western rites in the seventh century; but from the thirteenth onwards the missals prescribe them for the celebrant to say while the choir are singing the introit, Gloria, gradual and creed, in fact at every moment the liturgical action leaves him free. It is in this period too that we begin to find private prayers of the same kind inserted between the old public Agnus Dei and the priest’s communion. Let us be quite clear as to the point. Piety and edification may take many forms, and modern eucharistic piety still feels this particular mediaeval form to be entirely good and natural. But it is legitimate to point out the difference between this and the piety of the primitive church, for which the corporate liturgy itself formed the substance of devotion, and the corporate action its expression. In the middle ages it begins to be the supplementary prayers and the private emotions which take their place in this respect.

It is true that in the middle ages these ‘devotions’ are still only something which accompanies the liturgy, which continues uninterrupted by them, as it were, in the center of the field. In the sixteenth century a further stage is reached in practice. Either, as in a good deal of Latin devotion, the text of the liturgy is ignored altogether for purposes of devotion, and ‘methods of hearing mass’ entirely composed of these supplementary prayers are put together for the benefit of all but the celebrant. Or else, as in the sixteenth century Anglican rites, the supplementary devotions invade the liturgical action and become formal parts of it — even main parts of it, which break up the old apostolic and primitive action by an elaborate commentary of prescribed subjective repercussions which it is thought desirable that it should have on those present. The Anglican exhortations, the confession with its highly emotional language, the comfortable words and prayer of access — all of these are thoroughly in line with the piety of the mediaeval apologiae, and echo their language. The only change is that they are no longer private and supplementary prayers, but public and prescribed, and have been made a part of the liturgical action itself. Coming where they do, as a lengthy interpolation between the offertory and the consecration (which primitive Christian thought had seen as two parts of a single action so closely united that a single word prospherein would cover them both) these devotions would have been incomprehensible to the pre-Nicene church. That church never developed anything comparable because it understood the eucharist as being for all who took part in it an action — ‘Do this’ — not the experience of an action. Even when the post-Nicene church began to develop something of this kind, it placed them at the obvious point for such a development — before the communion. The mediaeval church on the contrary would have understood easily enough what Cranmer intended, even though it was itself still too loyal to primitive forms to bring about such an upheaval of the Shape of the Liturgy as this long interpolation involves.

The additional devotions after the liturgy which correspond to the ‘preparation’ before it — the ‘last gospel’ (John 1:1-14) and thanksgiving office (Benedicite, Ps. 150, etc.) — have much the same history. They begin as private devotions in the eleventh-twelfth century and have become a prescribed appendage to the rite by the sixteenth. They are part of that ceaseless process of accumulating ‘devotional extras’ around the essential liturgical action, which is the special mark of the piety of the ‘third stratum.’ In these cases its result is obviously edifying and good; in some others (e.g. the interlarding of the rite with private apologiae of the priest) it is difficult not to see in its manifestations only a false emphasis on inessentials.

The main lines of all the Eastern traditions had been reached before the end of the fourth century, and after this the process in all of them is no more than one of adjustment and development of detail. No new principle arose in the fifth century, as it did in the West, to give a new turn to liturgical development. After the sixth century the process resembles to this extent that which we shall find in the West, in that it is one of approximation between all the Eastern rites, as there is after this date an approximation between all the Western rites. And in both cases the basis chosen is the rite of the ‘holy city,’ Jerusalem in the East, Rome in the West. But in each case it is the rite of the holy city as modified in the dominant political center, Byzantium in the East, the Frankish homelands of Gaul and the Rhineland in the West.

But there the resemblance between the Eastern and Western process of development ends, for the methods pursued in the two halves of Christendom were different. The comparative freedom in which the churches were left to achieve the process in the West results in a real synthesis, in which the old local rites each contribute a good deal to the final result, and lose themselves in it. The methods employed in the East were different, consisting in political pressure and compulsory Byzantinisation. Not only did the Byzantine rite itself assimilate little or nothing from other sources after the sixth century, but the attempt was made to enforce its local development verbally and identically on all the churches which the emperor could reach. The legacy of Byzantine bureaucracy was too bitter for such tactics ever to succeed. The dissidents retained their liturgical independence of Constantinople by remaining outside the pale of orthodoxy.

But though the direct attempt of Constantinople to enforce its own liturgy failed entirely to bring about liturgical uniformity in the East, the general tendency, of which this was only a political perversion, to adopt a Syrian liturgy of the Jerusalem-Antioch type, has since operated throughout the East by voluntary ‘Serialization.’ The Egyptian monophysite version of S. Mark was heavily revised with borrowings from S. James and S. Basil in the fifth or sixth century. Later (? in the ninth century) it was replaced altogether, except on the Friday before Palm Sunday, by two alternative Syrian liturgies, a version of S. Basil (older than the present Byzantine text in some respects), and a liturgy addressed to the Son ascribed for some reason to S. Gregory. (There is no reason to suppose it has anything to do with him.) So the tradition which had come down at Alexandria from the apostolic age through Athanasius and Cyril was laid aside at Alexandria by the Copts. The Greeks after heavily Byzantinising it for a while, abandoned it altogether at the end of the twelfth century in obedience to Balsamon. Only the three dioceses of Uniat Copts now use S. Mark (or S. Cyril as they call it) even once a year.

The East Syrian rite of Addai and Mari has likewise acquired a considerable number of Antioch-Jerusalem characteristics at various times since the fifth century; and two alternative liturgies of the ordinary Antiochene type, ascribed to Nestorius and Theodore of Mopsuestia, have been brought into use. The Armenian rite has been affected both by the Byzantine version of the Syrian rite and by the Syrian rite of S.James itself, to the extent of becoming practically a rite of the Syrian type; though it s till retains a few interesting native features, and some Latin borrowings it picked up in Crusading times.

The Byzantine rite itself, clearly of Antiochene-Syrian derivation, continued to develop along its own lines down to the seventh century and did not become absolutely rigid until the ninth century. (Two complete revisions of the lectionary, for instance, can be traced in the seventh and the eighth centuries, none since). After that date only continual minor verbal changes in the prayers of the liturgy, and the accumulation of supplementary devotions during and before the preparation of the elements, can be traced. It is now used with only the slightest verbal differences throughout the orthodox world in a variety of translations; and once there ceased to be a Byzantine emperor looming behind it, its prayers and ceremonies and customs (e.g. the ikonostasion) have increasingly affected the rites and churches of the dissidents, especially in modern times.

We in the West are accustomed to speak of the ‘unchanging East’ and its ‘immemorial rites.’ It is as well to be clear that this is a state of things which only begins in the seventh century. Before that date the East had shewn more tendency to innovate in the liturgy than the West, particularly in the fourth and fifth centuries; and its rites, if they shew fewer signs of later development than those of the West, underwent much more drastic changes in that period than has been generally realized. What caused them to cease to develop in the seventh and to grow rigid after the ninth century is a matter for discussion. It is worth noting that this rather sudden ossification is a phenomenon which is found at about the same period in the whole artistic and mental life of the world that looked to Constantinople. But so far as the liturgy is concerned I believe that the use of the term ‘arrested development’ is unjust and untrue. It is only a case of ‘completed development.’ Without some fresh principle, such as the effect of the calendar on the prayers gave to the Western rites, the Eastern rites simply had no further possibilities of growth along their own lines. They were complete and satisfying expressions of the eucharistic action and its meaning according to the tradition of the churches which used them. There was nothing more to be said, nothing to be added. And into the closed world of Byzantium no really fresh impulse ever came after the sixth century. The Byzantine state had exhausted its own traditions by the ninth century, and then became mummified and finally disappeared. The Byzantine church survived it because it is the church, though the Phanar, ‘the royal church’ of Constantinople itself, has done little since to make that survival either fruitful or dignified. Orthodoxy is a far greater and more Christian thing than Byzantinism — rich in faith and holiness and above all in martyrs. Until this last twenty years it was still possible (though unfair) to call it a ‘sleeping church.’ But that sleep began not with the rule of the Turks in 1453, but in the ninth century, perhaps even earlier, in the sixth after Justinian. It will be fascinating to see what it makes of its magnificent patristic heritage in the modern world when it has been everywhere set free from its old entanglement with autocracy. One thing it will assuredly keep is the Byzantine rite by which all orthodoxy worships, and has saved itself from extinction by worshipping. This is the joint creation of Greek Christian theology and the old Hellenic poetic spirit, working together on a Syrian rite. Along with the Digest of Justinian it is the greatest legacy of Byzantine thought to the world.

The Western development is more complicated and diverse and continued for much longer. It will occupy the remainder of this chapter, and can most conveniently be set out by following up separately the various regional developments which come to their synthesis in the tenth century, and then continuing from that. But there are certain essential general observations which must be borne in mind all through, if we are to understand the matter.

The importance and interest of the special developments of the Gallican and Mozarabic rites have been much obscured in modern study. This is due partly to the fact that they have been for so many centuries virtually museum pieces, and it is correspondingly difficult to enter into their particular spirit. Partly also it is due to less excusable mistakings of their history and significance, the most serious of which is the persistent attempt to find for them a non-Western origin. These rites certainly contain Eastern elements (like the Aios and the Kyries and the Sanctus), just as the Roman rite contains Eastern elements (like the Kyries and the Sanctus and the Agnus Dei), and for the same reason — the deliberate piecemeal borrowing, now of one item, now of another, from Eastern and especially from Syrian sources. The Gallican and Mozarabic rites contain rather more of these items than the Roman only because Rome rather less readily admitted innovations from any source. But in all the Western rites these Eastern borrowings are relatively late and of superficial importance, matters of decoration rather than of substance. Structurally and in their fundamental spirit and origin these French and Spanish rites are as Western as any in Italy. Such structural differences as they exhibit from the Roman rite are due to slightly different arrangements of those lesser prayers of the ‘second stratum’, which only began to be introduced one by one into any of the Western rites about or after a.d. 400.

The question has often been debated as to the relation of these rites with those of Rome and Africa. Attempts have been made to shew that Africa used the ‘Gallican’ rite, or alternatively that it used the Roman. It has been held that the so-called ‘Gallican’ rite is really the original form of the Roman, faithfully preserved in the provinces when the mother-church (secretly and without record) turned its own rite upside down; or alternatively, that the churches of France and Spain originally used the pure Roman rite and that the whole of the Gallican and Mozarabic liturgical development in a novel and rootless local experiment of the dark ages. I can only say that this whole way of regarding the matter has come to seem to me not only mistaken but perversely unhistorical. And I suspect that it is not unconnected (however unconsciously) with partisan positions, for and against, on the modern problem of ‘Rome.’ In reality it is wholly unwarrantable to read back into the fifth and sixth centuries — or for the matter of that with any rigour into the seventh or even the eighth centuries, though the conception was developing about then — anything like the modern conception of ‘rites’ as defined and separate entities, ranged alongside one another in conscious difference and even in rivalry. Who is going to tell us whether the compilers of the Bobbio Missal or the Missale Francorum on the one hand, or the various Frankish ‘Gelasian’ missals on the other, supposed their books with their heterogeneous contents to be books of the ‘Gallican’ or the ‘Roman’ rites? Even with modern scientific methods of classification it is difficult, sometimes impossible, to decide; and what is quite clear is that the compilers themselves never even asked themselves the question. In the fifth and even sixth century, as in the fourth, there were still no ‘rites’ in our modern sense, but only ‘the liturgy,’ which everyone knew to be the same thing everywhere. Every local church had its own traditional way of doing it, which it was free to revise or augment or improve as it saw fit, from its own inventions or by borrowings from elsewhere. There were tentative efforts after local uniformity, like those of the Councils of Milevis and Vaison; but they were still occasioned by local circumstances, and limited and temporary in their real effect on what went on in practice at the altars in the churches. In every church contemporary fashions and novelties had their own attractions in each generation. Local tradition still played a preponderating part. In the long run racial temperament and characteristics (rather than geographical distinctions) made their different and immensely powerful influences felt on the wording of prayers and above all on the character of devotion and rites. In the circumstances this was inevitable; there were as yet no artificial national unities in the West, and Europe was in the melting pot. We know little enough about the African rites. But to an impartial view even the scanty evidence available indicates that they were neither ‘Roman’ nor ‘Gallican’ but African — the local development of the pre-Nicene African tradition, enriched by borrowings from other churches, not only Western but Eastern, but the whole moulded by the mind and spirit of the African local churches. The passage from the African sixth century prayer cited by Fulgentius (p. 297) indicates that it was not variable like the contemporary French and Spanish prayers. But it certainly is not ‘Roman’ any more than it is ‘Gallican,’ though it is quite easily recognizable as Western.’

And it was the same elsewhere. All the Western rites have their roots in the old pre-Nicene tradition, which as regards the Shape of the Liturgy was oecumenically the same. As regards the contents of the prayer the Western rites as a group have preserved the old conceptions of the eucharist more faithfully in some things than those of the East, which underwent more radical changes during the fourth century. Certain peculiarities common to the whole of the West (e.g. the ‘naming’ in connection with the offertory) make their appearance in the fourth century, and grow into real distinctions from the Eastern rites during the fifth and sixth centuries. (This is partly the result of different innovations being made simultaneously in the East.) All this is a consequence of the need for adapting the eucharist to a public worship. Most important of all for the future, the new Western principle of varying the prayers according to the calendar makes its appearance in the fifth century and is applied by the various Western churches in rather different ways, or perhaps it is truer to say, to a varying extent. In the course of the sixth-seventh century, when political confusion is great and intercourse between the Western churches much interrupted, these local Western differences in the application of a common principle harden into real distinctions, obvious to all and disconcerting to some minds, e.g. to that of S. Augustine of Canterbury (Cf. p. 576). The Roman ‘rite,’ the Milanese and Beneventan ‘rites,’ the Gallican ‘rite,’ the Mozarabic ‘rite,’ in our modern sense, are all substantially products of this period — it might even be said of the single sixth century. But in a.d. 600 men were not yet conscious of them as separate things, but still thought of them rather as different ways of doing the same thing. Each is the outcome of a local tradition and a local population living a local history; each is subject to particular influences from outside, as well as to local developments, working diversely upon the roughly similar basis all had inherited from the fixed rites of the fourth century, under the new influence of the ecclesiastical year and the calendar.

Viewed thus, as the native and characteristic products of the French and Spanish churches of the fifth and sixth centuries from their old liturgical tradition, the Gallican and Mozarabic rites come into their own, by coming into real life. They are the living response of French and Spanish Christianity to the sordid and desperate times when Europe had collapsed and civilization was struggling for a tolerable existence and the Faith had somehow to redeem to Christian goodness whole populations of uncouth and violent men and women. As such these rites have an exciting interest. And it is possible, I think, to shew that though they did not formally persist much beyond the dark ages which gave them this particular form, they yet handed on a permanent element to that synthesis of Western liturgy which is the slow work of the seventh to the tenth centuries.

The outstanding peculiarity of these rites is their treatment of the eucharistic prayer, in which, except for the text of the sanctus and the paragraph containing the narrative of the institution, the whole eucharistic prayer is varied, or ‘proper,’ on every liturgical occasion. Both in France and Spain this prayer consists always of five paragraphs:

1. The Contestation Illatio or Immolatio, or as we should say, ‘preface.’
2. The Sanctus, sung by the people.
3. A short paragraph, linking sanctus and consecration, known as the Post-Sanctus.
4. The institution-narrative, said in silence and known therefore as Mysterium or Secreta.
5. A prayer for the communicants, later changed to one for the offerings, as communions became infrequent, known as Post-Mysterium, Post-Seereta, or Post-Pridie.

Let us take examples. Here is the prayer of the eighth century Missale Gothicum (French) for the feast of the Epiphany:

Contestatio: ‘It is truly meet and right, just and right, that we should always and everywhere give thanks unto Thee, O Lord holy, Father almighty, everlasting God; Who didst lift up Thy voice unto us from heaven above Jordan’s banks like the sound of thunder: to point out the Saviour of the world, and shew Thyself the Father of the eternal Light, Thou didst open the heavens and bless the air and purify the waters, and shew Thine only Son by the Dove of the Holy Ghost. On this day the waters received Thy blessing and took away our curse; that they might offer to the faithful the washing away of all sins, and by regeneration make sons of God unto life eternal of those whom fleshly birth had brought forth to life in time. For those on whom death had laid hold by disobedience, life eternal recapturing them from death recalls to the heavenly realm. Wherefore with rightful exultation we join to the praises of the Angels our voices as we worship Thy glory in this wonderful sacrament on this day’s feast and offer unto Thee the sacrifice of praise for the Epiphany of Jesus Christ our Lord and for the source of our own calling unto Thee (i.e. baptism) through Him our Lord, through Whom the Angels praise, the Dominations adore, the Powers fear Thy majesty. The heavens and the powers thereof and the blessed Seraphim in common exultation tell Thy praises. With whom we pray Thee bid that our voices also be admitted, with suppliant praises saying:

‘Holy, holy, holy, etc.

Post-Sanctus: ‘Truly holy, truly blessed is our Lord Jesus Christ Thy Son, Who in token of His heavenly birth bestowed upon the world this day these wonders of His majesty: that He showed the worshipful star to the Wise Men, and after the passing of years turned water into wine, and by His own baptism hallowed Jordan’s flood; even Jesus Christ our Lord:

Mysterium: ‘Who on the day before He suffered ... (There follows the institution).

Post-Mysterium: ‘O Lord, we pray Thee, look with favour on these sacrifices before Thee; wherein no more gold and frankincense and myrrh is offered, but that which by these same gifts is declared, is (now) offered, sacrificed and taken. Through Jesus Christ our Lord Thy Son: Who with Thee and the Holy Ghost etc.’

This prayer illustrates a fairly common occurrence in the French prayers, the working in of prayers from the Roman rite into Gallican masses, as a rule in a rather different connection. The end of this Gallican preface for the Epiphany (from ‘through Whom the Angels praise . . .’) is taken from the Roman ‘common’ preface (not that ‘proper’ to the Epiphany); and the Roman offertory prayer for the Epiphany has been used for the Gallican post-mysterium.

Here, again, is a ‘pure’ Gallican prayer for use on any ‘green’ Sunday from the seventh century Masses of Mone, the oldest Gallican collection extant:

Contestatio: ‘It is meet and right that we should ever give thanks unto Thee, O God in Trinity, Whose power created us by Thy Word, and deservedly condemned our offences, Whose love delivered us by Thy Son, and called us to heaven by baptism and repentance: Unto Whom (all Angels and Archangels deservedly give ceaseless praises saying:)

‘Holy, etc.

Post-Sanctus: ‘0 God Who wiliest that we should not only offer to Thee the hymn but also the deservings of heavenly spirits, and should have no less the holy offices than the songs of the Angels: Grant that we who in setting forth Thy praises take to ourselves the united strains of the heavenly Powers, may also by amending our evil ways take to ourselves the love of the heavenly life, now that we are about to say those words of our Lord Jesus Christ which He left us for the memorial of His passion: through Jesus Christ our Lord ...

Mysterium: ‘Who on the day before ...

Post-Secreta: ‘O God of Abraham, God of Isaac, God of Jacob, God and Father of our Lord Jesus Christ, do Thou mercifully smiling down from heaven receive this our sacrifice with most indulgent love. May there descend, O Lord, the fullness of Thy majesty, Godhead, piety, power, blessing and glory upon this bread and upon this cup: and may it be unto us the eucharist Christ ordained by the transformation of (i.e. into) the Body and Blood of the Lord: that whosoever among us, and howsoever often, shall partake of this bread and this cup, we may take unto ourselves a memorial (monumentum) of faith, sincerity of love, and untroubled hope of resurrection and unending immortality in the Name of the Holy Ghost Who proceedeth from Thee and Thy Son in the communion of all saints, in the remission of all our sins. We believe, O Lord, that Thou wilt grant these things which we ask with unwavering faith. Through.’

It should be said that the compiler has carried the principle of variability to the length of equipping this prayer with an alternative contestatio besides the one here given — a variation of a variation — though this is not uncommon in the Gallican books. The frequent incoherence of the Gallican prayers is illustrated by the post-secreta here.

Finally here is the eucharistic prayer of the Spanish rite for the feast of S James (later the patron of Spain) which in the ninth century was kept on December 30th, the day after S. John:

Illatio: ‘It is meet and right that we should always give thanks unto Thee, O Lord holy, Father eternal, everlasting God, through Jesus Christ, Thy Son, our Lord: in Whose Name Thy chosen servant James healed the impotent man that cried unto him when he was being dragged to death, and by this miracle so moved the heart of him who mocked him that he brought him to attain to the glory of martyrdom when he had been instructed in the mysteries of the faith. So James himself fell slain by beheading for the confession of Thy Son: attaining in peace unto Him for Whom he bore this death. For He is Thy only-begotten Son Who gave His life as a ransom for many. Through Whom, O God the Father, do Thou bid that our sins be forgiven. Unto Whom all Angels and Archangels deservedly give ceaseless praises saying:

‘Holy, etc.

Post-Sanctus: ‘Truly holy, truly blessed is our Lord Jesus Christ the Son, Whom James leaving Zebedee his father so followed loving Him most dearly as to be chosen unto life, clean in conscience and approved in doctrine: at the last so commending his wisdom by his works that he died by beheading for Him Whom he knew had laid down His own life for himself and for all men.

‘Mysterium: Even Jesus Christ our Lord Who...

Post-Pridie: ‘0 God, bow down our necks under Thy yoke: that we may so bear Thy burden which is light unto them that love Thee with all desirable, devotion, as James Thine Apostle was joyfully dragged to execution with a rope around his neck; that sanctifying these things which we offer unto Thee, Thou wouldst bless us by the partaking of this Host (or Victim). Through...’

It is obvious that the Mozarabic and Gallican rites are, as regards the eucharistic prayer, variants of a single rite — scarcely even that, for the same technical terms, liturgical tags and phrases, even the same formulae, recur constantly in both. The distinction between them comes in the addition of two prayers of the ‘second stratum’, the ‘collect’ and ‘thanksgiving,’ which the Spanish churches were behind the French in adopting. (We may probably see in this a result of the more direct contacts of the French with the Italian churches in the later sixth century). But as regards the eucharistic prayer the Mozarabic and Gallican rites may be treated as being a single collection of variable prayers.

Nor, structurally, does there seem to be much difficulty in tracing the origin of this form of canon. It goes back plainly enough to the general fourth century fixed type of Western prayer, as revealed e.g. in Mai’s Italian prayers (p. 540). The preface and sanctus have replaced the ‘thanksgiving series,’ with an allusion to the liturgical commemoration of the day in the place of the old general ‘thanksgiving’ for the redeeming work of Christ. But the opening is the same, and most Gallican contestationes (like most Roman prefaces) are careful to retain at some point the per Quem (‘through Whom’), which is a notable feature of the Western ‘thanksgiving series’ as early as the prayer of Hippolytus. The post-sanctus is still the precise equivalent of ‘the link’ (Hippolytus e) between the ‘thanksgiving series’ and the institution narrative. But now it links the inserted sanctus with the institution. This latter is followed by a prayer for the communicants of precisely the same general type as that in Hippolytus (k) and the Supplices Te of the Roman canon. All that is missing from the Hippolytan outline is the anamnesis paragraph (h). But as we have seen, all the evidence suggests that this was still a local Roman peculiarity in the third and fourth centuries.

If we look back to the eucharistic prayer cited by S. Ambrose in de Sacramentis as the contemporary Milanese and Roman canon, we find that after the laudes (= ‘thanksgiving series’) and the asking of ‘prayers for kings, for the people’ (= the ‘Names’) it runs as follows:

1. ‘Make for us this oblation approved, ratified, reasonable, acceptable, seeing that it is the figure of the Body and Blood of our Lord Jesus Christ:

2. ‘Who the day before He suffered ... (there follows the institution).

§ ‘Therefore making the anamnesis of His most glorious passion and resurrection from the dead and ascension into heaven,

§ ‘we offer to Thee this spotless offering, reasonable offering, unbloody offering, this holy cup and bread of eternal life:

3. ‘And we ask and pray that Thou wouldst receive this oblation at Thine altar on high by the hands of Thine angels, as Thou didst vouchsafe to receive the offerings of Thy righteous servant Abel, and the sacrifice of our patriarch Abraham, and that which Thine high-priest Melchisedech offered unto Thee. (That as many of us as shall receive by this partaking of the altar the most holy Body and Blood of Thy Son may be filled with all heavenly benediction and grace).’

The paragraphs marked § are already present in substance in the local Roman prayer of Hippolytus c. a.d. 200. For the rest, it seems easy to recognise in 1, 2, 3 ‘the link,’ the institution and the prayer for the communicants of the Gallican post-sanctus, mysterium and post-mysterium, and of Hippolytus e, f and k. The main differences between the Franco-Spanish and Italian developments are 1. That the Italian prayers place ‘the Naming’ in the second paragraph of their eucharistic prayer (this is probably a fourth century innovation, Cf. p. 501) whereas the Franco-Spanish rites place it at the offertory (probably the original Western position). 2. That the Roman prayer (if not other Italian prayers also) retains an old pre-Nicene peculiarity in inserting the anamnesis clauses (§ §) between the institution and prayer for the communicants. But apart from this all the Western prayers have the same structure.

When the evidence is set out, no one could easily suppose that the Gallican eucharistic prayers as they stand represent any very ancient survival. They are too completely affected in their contents by the sanctus and the influence of the calendar, neither of which, as we have seen, make their appearance in the Western rites till the fifth century. The preface and the ‘link’ are dominated by both these influences, and even the prayer for the communicants is frequently overwhelmed by the allusion to the day (c/. the mass of S. James on p. 555). Their fidelity to tradition consists in arranging their new contents on the old Western scheme. Only the institution narrative itself, now regarded as too sacred and too important as the consecration formula to be lightly varied, has survived unchanged from before the acceptance of the new fashion of variability. For the rest, the very fact that a fresh composition had to be found for every liturgical day in the year prevents us from hoping to discover any surviving trace in the Gallican prayers of the actual wording used in Gaul and Spain in the fourth century before variability came in. At the best a church could only keep its traditional prayer as one variant, for days which had no particular liturgical associations, e.g. ‘green’ Sundays. But all the Gallican collections extant provide a whole set of alternatives for these, none of which fail to conform to the later Gallican type. The French and Spanish local eucharistic prayers of the fourth century and earlier seem to have transmitted only their structure, not their wording, to their successors.

The date when the Spanish and French eucharistic prayers first became variable with the calendar cannot be satisfactorily fixed. It is clear that when Pope Vigilius in answer to the enquiries of Bishop Profuturus of Braga (Portugal) described the fixed Roman canon with variable insertions on certain great feasts in a.d. 538, he was already aware of a difference of practice in this between Rome and the Spanish churches, though he does not press the point. S. Isidore of Seville attributes the composition of ‘prayers well adapted for various feasts and masses in an elegant style and lucid phrasing’ to Peter, bishop of Lerida c. A.D. 500. Though this does not specifically refer to the eucharistic prayer as such, Isidore, who used the Mozarabic rite, would doubtless not have considered a mass which did not include a complete ‘proper’ eucharistic prayer ‘well adapted’ for a feast. There is no reason to suppose Peter was the first author of Spanish variable prayers, but the names of the earlier authors have not been recorded.

In Gaul we seem to have such an earlier record. Musaeus, a presbyter of Marseilles (d. c. a.d. 460) is said ‘to have compiled at the request of his bishop Saint Venerius lections from the holy scriptures suitable for the feast days throughout the year; and also responsories from the psalms and versicles and responses (capitula) fitting the seasons and lections; which book is so far considered a necessity by the lector in the church, that it relieves him of all fuss and worry and does away with delay, and at the same time instructs the people and gives fitting honour to the feast. And he also composed and dedicated to the bishop S. Eustace, the successor of the aforesaid man of God, a remarkable and fairly long Book of Sacraments, divided into sections for the sequence of offices and seasons, and for the text of the lections and the arrangement and chant of the psalms; but displaying his usual earnestness both in prayer to God and the acknowledgment of His goodness.’ Musaeus had some reputation as an exegete, but this part of his work was clearly liturgical. The book composed for bishop Venerius (d. c. a.d. 452) seems to be for the office, and it has been questioned whether the volumen sacramentorum dedicated to his successor was not a book of instructions or homilies rather than a ‘sacramentary.’ It seems sufficient answer to point out that Liber Sacramentorum is the official heading of the Gelasian Sacramentary compiled c. a.d. 475-510. Gennadius also tells us that Voconius, bishop of Castellanum in Morocco c. a.d. 460 wrote another volumen sacramentorum, probably an African sacramentary, but there is no indication in this case that the prayers were arranged according to ‘the sequence of the offices and seasons’ (and presumably varied with them) as in the work of Musaeus.

Eustace seems to have become bishop c. a.d. 452 and Musaeus died c. a.d. 460. Once again, as in the case of the Milanese ‘proper’ prefaces, we are pointed to the period about or soon after a.d. 450 as that of the introduction of variable prayers in the Western rites. Musaeus may not have been the first author of such prayers in Gaul, but he is the first to be recorded. And Gennadius writing c. a.d. 495 gives a fairly full account of even the lesser ecclesiastical writers of southern Gaul in the fifth century. It seems hardly likely that he would have passed over ecclesiastics who had made any considerable name for themselves as liturgical authors in a new genre during this period. Musaeus need not be regarded as personally responsible for the invention of variable prayers in general, or even of only those of the Gallican rite. The idea seems to be too widespread too suddenly in the latter half of the fifth century to have had any single inventor. Probably it was in the air, a consequence of the new ecclesiastical year which had now dominated the whole celebration of the liturgy for more than a generation as a fixed and accepted institution of church life. The new fashion, coming in sporadically and haphazard, may well have been the occasion for Musaeus’ orderly and systematic compilation much more than the consequence of it, even at Marseilles itself. And other South French churches doubtless made their own terms with it at about the same time, though they could not command the services of a well-known scholar to refurbish their liturgical traditions, and their obscure and tentative compilations have in consequence left no trace. Even Musaeus is not said to have written variable eucharistic prayers, though the word contestandi inevitably recalls the Gallican term contestatio for the variable preface. In any case once the first paragraph of the prayer had been made to vary with the liturgical feast, the idea of varying other paragraphs of the prayer in accordance with it need not have been long in presenting itself to someone. The admission of merely alternative texts (not dependent on the calendar) of the whole prayer earlier in the century had already undermined the fourth century idea of a single fixed eucharistic prayer unvarying on all occasions. All things considered, I think we may safely date the general acceptance in France and Spain of variable eucharistic prayers in the latter half of the fifth century, with perhaps a period of preliminary and tentative beginnings in the ten or twenty years before that.

It hardly admits of question, from the mere identity of structure, that the Gallican and Mozarabic rites spring from a single source and are indeed only a single rite. Whether it originated in Gaul or Spain there are no decisive means of telling (though my own guess would be in favour of Gaul). The oldest surviving French MS. (the Reichenau palimpsest containing the Masses of Mone) is dated c. a.d. 650. The oldest Spanish MSS. are only of the ninth century, but what is recognisably the Mozarabic rite is described by S. Isidore of Seville in his de Ecclesiasticis Officiis in the early seventh century. Spanish tradition usually ascribed the rite itself to his compilation. But whatever lies behind the tradition, mention of Peter of Lerida as the author of some of the prayers more than a century earlier shews that Isidore’s work can have been no more than a revision and reorganisation, akin to that carried out by S. Gregory in the Roman rite in the same period.

The great mutability of the eucharistic prayer in these rites was against the building up of any very stable tradition. When the laity expected to hear an entirely different set of prayers every time they went to church, they were not likely even to know whether this year’s set, e.g. on Ascension Day, was the same as last year’s, since only the celebrant had a book. The permanent tendency of the clergy to innovate in the text of the liturgy was thus released from the usual check of the layman’s attachment to a familiar form, except so far as concerned the structure of the prayer and certain obvious cues, before the sanctus and the concluding ‘Amen.’ Thus though we can be sure that the special characteristics of this type of prayer were in general accepted by the churches of Spain and Southern Gaul by c. A.D. 500, it is not safe to take it for granted with our present knowledge that the texts which we have are necessarily much older than the extant MSS. which contain them. None of the seventh century texts of the Masses of Mone are found again in the eighth century Gallican books. Elipandus, bishop of Toledo in a.d. 794 cites from the masses found in the ninth century Mozarabic Liber Sacramentorum for such important days in the calendar as Maundy Thursday, Ascension Day and others, and he ascribes each mass by name to its author. All those he mentions are bishops of Toledo after a.d. 650. He is writing officially on behalf of the whole Spanish episcopate to the bishops of Gaul. His statements can hardly be made at haphazard; and if the attributions of the authorship of the various masses had not then been certainly known at Toledo, it is strange that the prayers should have been fathered on comparatively recent writers, and not on Isidore or some other great name of the more remote past, for Elipandus is anxious to impress. Of course, these seventh century bishops’ ‘authorship’ may have consisted in no more than a mere revision of older work and the attachment to it of their own names. Yet little more than a superficial investigation of the Mozarabic and Gallican prayers is needed to shew that they come not only from many hands, but from more than one period of taste and latinity. Some may well be as old as the later fifth century (e.g. the Mozarabic masses for S. Martin) but others are undoubtedly from the ninth century, after the Moorish conquest. It may be that one day we shall be able to distinguish more easily than we can at present the earlier from the later in the main bulk of these prayers.

It remains to say something of their distribution and history. The Mozarabic rite was codified as the rite of the see of Toledo, whose archbishop is still ‘Primate of the Spains.’ But the ecclesiastical greatness of Toledo dates only from the conversion of the Visigothic kings from Arianism in the late sixth century; and it was only in a.d. 633 that its rite was made the standard for the whole of Spain and the Visigothic dominions in the South of Gaul. Previous to that the various provinces had tended to adopt the rite of the local metropolitan. No doubt most of these were of the Mozarabic type, and some of the prayers of the Toledan missal were undoubtedly drawn from these older provincial and local ‘propers’ (Cf. p. 380). But the national council of the independent Suevic kingdom of Galicia held at Braga in a.d. 565 had ordered the use of the Roman rite. The use of the Toledan Mozarabic rite was enforced in Galicia as a political measure by the Visigoths when they conquered it, and it thus became the national rite of Spain.

It remained such down to and after the moslem conquest in the early eighth century. In the eleventh century the fringe of independent Christian principalities in the North and West began to adopt the now general Western rite. This was partly under the impulsion of French monks from Cluny who were unaccustomed to the Mozarabic, partly because, engaged as they were on a perpetual crusade for the reconquest of their country, the Spanish princes and peoples themselves were more conscious of their own unity with the rest of the Christian West. The Mozarabic remained the rite of the Christians living under the yoke of the Caliphs of Cordova. But as the tide of Christian reconquest advanced during the middle ages, so, too, did the Roman rite, which had now become the badge of freedom. By the end of the fifteenth century the Mozarabic rite had all but died out, being used only in some of the parish churches of Toledo and occasionally in the cathedral, and in some scattered churches elsewhere on a few occasions in the year. It was rescued from extinction by Cardinal Ximenes in a.d. 1500, who provided for its continuance in a somewhat Romanced form in seven Toledan parish churches and a specially endowed and staffed chapel in the cathedral.

The question of the diffusion of the Gallican rite is more difficult. Every single extant liturgical MS. of the Gallican rite can be traced back either to Burgundy or the country to the south-west of it (the Narbonnaise and Acquitaine) i.e. either to the region of France most accessible to Visigothic Spain and in intimate relations with it, or to the actual original nucleus of the Visigothic state. This is a fact not to be lost sight of in considering whether Gaul or Spain is the birthplace of the rite; but too much should not be made of it since the evidence of Gregory of Tours makes it probable that in the sixth century this rite was used also at Tours, which lay outside the Visigothic sphere after a.d. 496.

The problem arises as to the rite used in the North and East of France. The earliest MS. which has reached us from the church of Paris is a copy of the Roman Gelasian Sacramentary written c. A.D. 700, probably at S. Denis. And there is no doubt that the use of the Roman rite, at all events in certain churches, goes back in the North and East to a period a good way before a.d. 700 and probably before 600. This region may have used something like the Gallican rites of the South of France before that date. But we have seen that the Gallican rites really only begin to grow up in the South in the later fifth century. It is conceivable, therefore, that the Roman Gelasian book was the first compilation of variable prayers to succeed the old fixed rites in the North There is no evidence either way. We have also seen that Ireland used a form of the Roman rite soon after A.D. 600 and perhaps earlier. The Anglo-Saxon churches did the same from the landing of Augustine (a.d. 596). The real sphere of the Gallican rite after a.d. 600 seems therefore to have been confined to the center and south of Gaul. Burundian missions had begun to carry the Gallican rite to South Germany in the seventh century, just as Augustine found the Burundian bishop Liudhard before him at Canterbury using the Gallican rite in the private chapel of Queen Bertha, who had been a Burgundian princess before she married the king of Kent. But the definitive conversion of both England and Germany was effected by missions using the Roman rite, and the Gallican never took root in either country.

In France itself it fell into great decay during the eighth century, though it held on in the South and South-West until the time of Charlemagne c. a.d. 800, who formally abolished it. It is possible, however, that it did not finally die out in scattered churches for another fifty years or so after his time. Thereafter it survives only in certain sporadic ceremonies continued in many French churches, and as a pervading influence in the Romano-French liturgical books which resulted from Charlemagne’s reform.

The Gallican rite as a rite had therefore an effective life of some 400 years, from the fifth century to the ninth. The Mozarabic rite lasted for another two centuries in Spain, and took another three or four to fade into the position of an isolated local peculiarity in a handful of churches. In each country their disappearance coincides with the transition from the barbarian centuries to the new Europe and the beginnings of the resurgence of civilisation.