?r 563^ 59- Shmona and Guria ask for time to pray: the executioner asks for their pardon and their prayers. They pray in silence. §61. Just before kneeling down for execution they utter a short exclamatory prayer (16 words). § 60. The martyrs beg the executioners to bury them decently, as none of their friends are present. In § 65 we learn that they placed the bodies side by side, but left them unburied. First Shmona (§§ 62, 63), then Guria (§ 64), kneels down and is beheaded with a single blow. § 67. Shmona’s daughter mentioned as among the first group that goes out to find the martyrs’ bodies. [p. 60] Sharbel refuses to drink the murderers’ wine, because (he says) “I desire to feel the saw with which ye saw me as well as the sword which ye pass over my neck”. [p. 60] Sharbel asks for time to pray. The prayer is given in full (48 words with allusions to the Gos¬ pel History). [p. 60] Sharbel is elaborately done to death with absurd tortures, “and when he was at the point to die, because the saw had near¬ ly reached his mouth, they smote him with the sword”. [pp. 60, 61 1 Sharbel’s sister Babai catches his blood on her skirts. The city sheriffs go off and tell the judge, who commands HISTORICITY OF THE MARTYRDOMS. 15 them to return and order the executioners to kill her. They do so with tortures. The Christians bury the [p. 61] Christian sympathi- bodies without molesta- sers ') steal away the tion. bodies of Sharbel and Babai secretly. Surely this comparison speaks for itself. But the story of the execution of Shmona and Guria is not only more probable than that of Sharbel and his sister; it is, I venture to say, intrinsically probable. At the time of the ‘Diocletian’ persecutions Edessa was already a pre¬ dominantly Christian city. It only produced three Mar¬ tyrs, but the popular sympathy was on their side. Indeed we may well imagine that the central Government would be far more cautious in opposing what was almost a national cult, than in dealing with the religion of a minority. It was therefore prudent to make the execu¬ tion private. The episode of the cart is told most natu¬ rally: the executioner was in a hurry to get to his destination, he wished to spare his prisoners, one of whom was aged and infirm and the other could not walk, so he commandeers a chance-met “Ekka”, for no doubt the Zenge was a country bullock-cart. The anxiety of Shmona and Guria for decent burial is natural, espe¬ cially in an Oriental Christian, but the puppet-martyr of the martyrdom-literature is for too exalted to think of such things. Finally, the mention of Shmona’s daughter 1) I venture to read with B. M. Add. 14645 (as in Habbib 38) instead of . 1 6 INTRODUCTION. at the end leads to nothing: it appears simply as a pathetic circumstance. The fact alleged is not improbable in itself, but if it were a mere invention we may re¬ member that the author had at least an even chance of falling into error, for he has already described Guria as “sanctified”, i.e. as a celebate ascetic. Is it not easier to believe that “Theophilus” escapes all the pitfalls, be¬ cause he is really describing a scene which he had witnessed ? But if important sections of the narrative of “Theo¬ philus” approve themselves as contemporary and authentic, then we are justified in regarding the whole story with a prejudice in its favour: it becomes reasonable to treat the details that are incredible or contradictory as later accretions or corruptions of an historical account, rather than as the natural features of a fictitious narrative. The Text. — It is agreed that the Syriac is the ori¬ ginal, and that the Armenian (V) and both v. Dobschiitz’s Greek texts ((6, (b2) are translations. Further we have only a single copy of the Syriac; in the case of Shmona and Guria we have to depend on a MS. of the 15th century, in the case of Habbib on one dated 936 A.D. ') But I am sure it would be a mistake to regard all the extra portions of the Syriac text as later corruption or am¬ plification. Von Dobschtitz himself admits in the Greek translations a tendency to curtailment (p. Xlll), and most appropriately refers elsewhere to Abbot Butler’s remark¬ able paper on apparently “conflate” readings (p.xxx, note). 1) It is B.M. Add. 14645, the same that contains the Acts of Thomas. The dating in v. Dobschtitz, pp. xi and 1, who says “13th or 14th cen¬ tury”, rests on a misapprehension. HISTORICITY OF THE MARTYRDOMS. \J Especially is it hazardous, in my opinion, to reject the unconventional perorations at the end of the two Acta (5. & G. §§ 68 — 70, H. § 40). On the other hand one or two clauses of the Syriac, as we have it, are no doubt interpolated. In 5. & G. § 1 the MS. adds K'Au’w.to to rd'paln ^lia, i.e. adds the later term for “cloistered nuns” to the ancient term for Christian ascetics. This is omitted in the versions and is no doubt a late insertion. Again, in § 32 the three months’ imprisonment of the Martyrs in a dungeon is given in all our authorities: it is only in the Syriac MS. that the clause “without eating or drinking” is added. This clause also is no doubt due to a later scribe rather than an early editor '). I venture to think that the mention of the year in § 69 is an addition of the same kind. The very construction of the sentence shews the words to be an intrusion, and the parallel passage in Habbib § 39 , which gives the month and day of the martyrdom, also leaves out the year 1 2). The recognition of this last interpolation is as impor¬ tant as that of the interpolation in § 32. The removal of the interpolated clause in § 32 takes away the only part of the sufferings of the saints which altogether pass credibility, and the removal of the date in § 69 takes away the one clause in Theophilus’s account of the 1) The same may be said of the words “and his head” in § 35, as v. Dobschiitz points out. 2) I give the passage in full, with the clause which I think an inter¬ polation marked in italics: “And I have written these things on the 20th of November, on a Sunday, five days after the crowning of these holy Martyrs year 618 of the Greeks , whose very murderers were calling them blessed”. Surely any words between “holy Martyrs” and “whose” must be interpolated. 3 i8 INTRODUCTION. execution which could not have been written in good faith. With this clause away; we may ask ourselves why we should not regard Shmona and Gnria as based upon a genuine account by Theophilus, in the same way that the Acts of the Apostles contains a genuine travel-diary ? I believe that the parallel goes further and that Theophilus wrote the whole of the Syriac text, apart from a few corruptions such as the glosses I have just mentioned. The Historical Situation in the Acts. — Professor v. Dobschiitz in summing up against the historicity of these Acts lays stress on the similarity of style between them and the Acts of Sharbel. It seems therefore worth while to draw attention to certain points in which Shmona & Guria and Habbib do not follow “common form”. There is certainly very little in the story of Sharbel or in that of Barsamya to detain the historian. Sharbel the chief priest of the pagans is suddenly con¬ verted by the courageous bishop Barsamya, who does not fail to make the appropriate reference to King Abgar, to Addai the apostle, and to his predecessor Palut. This conversion takes place in the most dramatic manner, but without apparent psychological reason, just at the time of the arrival of a command from Trajan to enforce the sacrifices. Sharbel, the new convert, quotes Scripture with the judge, who rails at him and afflicts him with impossibly severe tortures. The notaries write an account of all that has happened, and lay it up among the city archives !). i) That is to say, “so that the Christians can find them in due time!” The real meaning is to give an explanation why this imposing event with all its wealth of edifying detail was unknown to genuine tradition and had not influenced the Christian Kalendar. HISTORICITY OF THE MARTYRDOMS. 19 I venture to claim that the picture presented in the Acts of Shmona and Guria and of Habbib is very dif¬ ferent. In Sharbel the central government is all-powerful : nothing but the miraculous power of Christ can stand up against it. In the Diocletian Acts, on the other hand, an attentive reader can see some of the difficulties of the Governor. By the beginning of the 4th century A.D. the Christians must have been a very powerful party in Edessa. The converts during the past century had included some of the old royal family and the noble philosopher Bardaisan. Moreover Christianity had begun to spread widely in the neighbouring Empire of Persia. Hardly a generation later the secular Powers had taken sides: the Byzantine Empire was Christian, the Sasanian was anti-Christian. But at the beginning of the 4th cen¬ tury the Christians were merely regarded as bad citizens of both realms, and the authorities on the spot must have felt the danger of turning bad Romans into actual Persians '). In Nicomedia Diocletian could do very much as he liked. In Syria also there was no alternative to the Roman dominion. But Edessa lay near the frontier, where a disaffected population might prove dangerous. Lactantius tells us [Inst. v. 11) that he had heard provincial Governors boasting that their administration had not been stained with Christian blood. It appears to me that the Governor of the Edessene district did his utmost to be one of these. In the Note on the Date of the Martyrdoms I have given in detail the reasons 1) I cannot help feeling that when Shmona and Guria announce that they will do the. will of the King of Kings (§ 12) it must have had a very ugly sound to a Roman Governor east of the Euphrates. 20 INTRODUCTION. which seem to me quite clearly to shew that the mar¬ tyrdoms took place in 309 and 310. If this be so, it means that the Governor did not take any serious steps to interfere with the Christians until six years after the great persecution had been started in Nicomedia. It is possible that the mention of the year 618 of the Greeks (= A.D. 306/7), with which our present Syriac text of Shmona ayid Guria starts off, may be original after all, for it is not quite clear how long is supposed to be occupied by the events described in §§ 1 — 17, especially § 4. The Governor has had Shmona and Guria arrested with other Christians. Of these others, some were flogged and set at liberty, some were more seriously hurt before being let loose, others fined — the Syriac MS. here adds that “many others were killed”, but the clause has no equivalent in either Greek version or in the Armenian — and others the persecutors simply let go (§ 4). “So Guria and Shmona his companion remained alone in prison” (§ 5 init.), comforting and encouraging themselves by hearing of the steadfastness of the Chris¬ tian Martyrs — in other places ! Surely this is a very unconventional picture of the times of the “Diocletian Persecution”. Does it not shew that the Governor must have been doing his utmost to do nothing? At last Shmona and Guria are brought up for trial before him. In §§ 6 — 14 he tells them that they must obey the orders of the Government, and on their refusal he has them kept in custody (§ 15). Then the Governor is sent for to Antioch: no doubt his leniency has been remarked, and he comes back to Edessa with express orders to proceed to extremities at once. The same general situation is depicted in the Acts HISTORICITY OF THE MARTYRDOMS. 21 of Habbib. The ‘persecution’ is entirely official. “When many were being persecuted they cried out of their own free will ‘we are Christians’ and they were not afraid of the persecution, because those who were per¬ secuted were more numerous than those who persecuted them'’ (§ i). The Governor sends to arrest Habbib, who happers to be away at the time, so his mother and other inhabitants of his native place Tel-she ') are taken prisoners to Edessa. Habbib hears of this and comes to Edessa to give himself up. The old captain of the guard would dissuade him and tells him not to be anxious about his relations, “for”, he says, “no man will hurt them, but they will remain a few days in prison and the Governor will then dismiss them” (§ 9). The con¬ versation between Habbib and Theotecna may be ima¬ ginary, but the situation depicted is wholly unlike the conventional picture of the faithful confessors and the ruthless heathen Governor, of which Sharbel and Bar- samya are typical examples. When finally Habbib is brought before the Governor, the annoyance of the Governor with Habbib is made to start from the fact that he has given himself up willingly. The Governor would have been better pleased not to have caught him. It had evidently been no part of the Governor’s plan to make Christian martyrs ; he only wanted to “save the face” of the government. This was naturally misunderstood by the translators and they have left out § 13 a where it is most clearly expressed. The Tortures. — The evidence of Eusebius in his Martyrs of Palestine is enough to shew that severe 1) I. e. ‘Drymount’: algo mentioned in Wright CBM 648. 22 INTRODUCTION. tortures were really used against Christians during the great persecution. Those which Shmona and Guria and Habbib are said to have endured are similar to those described by Eusebius, and much milder and less fan¬ ciful than those in Sharbel and Barsamya . The really noteworthy feature about the tortures in Shmona and Guria is that the martyrs themselves feel them. Old Guria is stretched one day on a vertical rack (§ 28) and afterwards confined in a dark dungeon (§ 32). He remains firm indeed in his confession, but physically he is a broken man. For this reason they give him no more torture (§ 35). Shmona, a younger man, is further tor¬ tured by being hung up by one leg, with the result that the tendons of his knee were severely damaged, and he cannot walk. As a rule in this class of literature, the most elaborate tortures seem to produce no phy¬ siological reaction whatever. Alleged anachronisms of Christian organisation. — I have already noticed the evidence which shews that the term dairy atha (“cloistered nuns”) does not occur in the genuine text of S'. & G. § I, and that the only genuine term there used for Christian ascetics is bnatlt kyamd (lit. “daughters of the covenant”). I cannot regard this term as an anachronism. Notwithstanding all criticisms, I have seen nothing to change my belief that in very early times the Church in Mesopotamia required the Marcionite rule for full Church membership, and that the Baptized Christian was expected to give up both his family ties and his property ’). How long this rule lasted is a matter of dispute : in practice, as distinct 1) See Early Eastern Christianity , esp. pp. 129, 142. HISTORICITY OF THE MARTYRDOMS. 23 from ecclesiastical theory, I suppose it broke down soon after the fall of paganism and the adoption of Chris¬ tianity as the official religion of the Roman Empire. But I see no reason to suppose that in the 3rd century, while Christianity was still more or less an illicit religion, persons in Mesopotamia were admitted to baptism who intended to live a married life. And this is the state of things reflected in our Acts. There is nothing to shew that Shmona (who had a daughter) was even baptized, still less that his daughter had been born after he had been baptized. In such a state of things a word is needed for those who are sympathisers with the Christians, though they are still living in the world. They are not yet properly “Christians”, but they are not “pagans”. Accordingly they are called rtlr-inc' ’ nashe Q aim aye, i.e. per¬ sons living in the world”. The Bar kyama is the baptized Christian, the cAlmdyd is the Christian sympathiser, not yet a Catechumen. At a later period, when Bar kyama had become, with changed circumstances, narrowed down to denote a particular kind of monk, a similar term, cdlmandya, is used for “layman”, i.e. a baptized Christian not in Holy Orders. But Habbib 38 reflects an earlier state of things. The body of Habbib is carried by ’ nashe calmaye as well as by “brethren”, and these in turn are distinguished from Jews and Pagans. Method of Compilation. — At the end of the Syriac text of Habbib , after the peroration , we find added : “Now at the 27th question that the judge asked Habbib, he gave against him sentence of death by the burning of fire”. This covers Habbib §§ 14 — 34, and I cannot 24 INTRODUCTION. help wondering whether it was not all that Theophilus could get out of the official minutes of the trial, and that his rough notes have somehow got tacked on to the end of his finished composition *). The story of the three trials of Shmona and Guria reads to me more natural than that of Habbib and more as if it was based on definite information. No doubt it is not a mere transcript of shorthand notes, as the Acts of the Scillitan Martyrs seem to be, but I do not see why it should not be an expansion of such notes. I do not see why §§ 37 — 39 should not be essentially historical : the judge hears Shmona muttering something as he hangs in torment; he hopes it may be a promise to yield, and he tells the Notary to go and take his words down. That Shmona said anything so elaborate as § 37, 38 (not that what we read is really elaborate), is impro¬ bable. But the general sense is what we might expect. “God of Stephen the Martyr, keep my mouth from begging for mercy in this agony!” — that is the gist of what he says. The difference between this and the text of §§ 37, 38, is only the difference between the literary taste of the fourth century and that of the twentieth. The chronology of the opening paragraphs in both Martyrdoms seems hopelessly faulty : the paragraphs were probably drawn up in their present form half a century after the events, when the Acts were adapted for ecclesiastical use. By that time the chronology of the events of the great persecution had faded from popular memory, and Licinius was remembered as the 1) See further, pp. 180 — 182. HISTORICITY OY THE MARTYRDOMS. 25 adversary of Constantine and as the author of vexations anti-Christian regulations '). In any case, if we accept the story of the execution of Shmona and Guria as the genuine report of an eye-witness, written in good faith, it is the occasional and indirect chronological indications that we ought to start from, as they are less likely to be altered. The grain of history hidden in the chaff of Sharbel and Barsamya is no doubt small. It is quite conceivable that among other noble inhabitants of Edessa a heathen priest named Sharbel may have been converted about the middle of the 3rd century, and such a personage may very well have been put to death about the time of the Decian persecution. It is noteworthy that the tradition, such as it is, does not make Barsamya the Bishop to be a martyr. The date of our Acts of Sharbel appears to be earlier than the episcopate of Rabbula (411 — 435), for Sharbel quotes the apocryphal Third Epistle of Paul to the Corinthians1 2). This Epistle was included in S. Ephraim’s Commentary on the Pauline Epistles, but naturally finds no place in the Peshitta New Testament as revised by Rabbula. On the other hand Barsamya tells Sharbel (p. 43) that aHe who put on a body is God, Son of God (i.e. hoe U Qeov), Son of the ithutha of His Father, and Son of the Kyana of Him that begat Him 3 *). This 1) See Gorres, Kritische Untersuchungen u. d. Licinianische Christen - verfolgung passim. 2) See ASD, p. Vi (E- Tr-i P* 56-2i): “The scars of my body — that I may come to the resurrection from the dead”. This is from the end of the Epistle. 3) Cureton ASD 43. That the theological terms here used correspond to the Nicene phrase opoovoiov tw nocT-p/ is fully explained in the Ap- 4 26 Introduction. is distincly post-Nicene language, and in an Edessene document it indicates the latter half of the 4th century, possibly even some time in the last quarter. Yet there is one feature of the Acts of Sharbel which demands notice here, because of its marked difference from the Acts of the Diocletian Confessors. Not only is the Christianity of the Christians different in the two sets of documents; the Paganism of the pagans in dif¬ ferent also. Sharbel was the priest of “Bel” and “Nebo”, and his vestments include the Diadem (k'.icuj), which according to the Doctrine of Addai was worn by the heathen Abgars of Edessa and is still to be seen figured on their coins. But the worship in which the Diocletian Confessors refuse to join is the worship of “Zeus”. In Sharbel there is no word of Zeus, in Shmona and Guria and in Habbib there is no word of Bel and Nebo. Moreover in Sharbel the populace of Edessa is repre¬ sented as half pagan, half Christian : in the Acts of the Diocletian Confessors the city is overwhelmingly Chris¬ tian ‘), and the persecution of the Confessors is entirely the work of the distant and alien Graeco-Roman Em¬ peror. “This Zeus” is the alien God, whose worship in every market-place was inculcated by the Fifth Edict of the Persecution. It should be noticed that it is not only a question of the names given to the Gods of heathendom. Bel and Nebo, so far as they do not signify the planets, may very likely come from the Old Testament. But pendix to Bethune-Baker’s Nestorius , p. 217, to which the curious reader is referred. 1) See especially Habbib § 1 (end). HISTORICITY OF THE MARTYRDOMS. 2J when the judge says to Sharbel “Renounce not the Gods whom thou once confessedst”, and Sharbel replies “Spare me again from saying that there be Gods and Powers and Fates and Nativities” l), are we not moving in the circle of ideas in which the Bardesanian Dialogue on Fate moves? There is nothing at all like it in the Diocletian Acts. I do not mean to try and rehabilitate on these grounds the historicity of the Acts of Sharbel. What I claim is that the Acts of Sharbel and the Acts of the Diocletian Confessors cannot both be fictitious and also by the same hand. It is not likely that a historical novelist would on the one hand so delicately distinguish succes¬ sive stages in Edessene paganism, while at the same time he makes Sharbel in the distant past (but not Shmona, Guria and Habbib in the nearer past) use the shibboleths of Athanasian controversy. Whether the existing Syriac MS. of Shmona and Guria has more accurately preserved than the versions the terms in which the Martyrs speak of Christ is a difficult question, but in neither case are they made to use the Athanasian catchwords or the Trinitarian formulae. Those who main¬ tain the literary connexion between the Acts of Sharbel and the Acts of our Confessors, and yet consider these latter Acts fictitious, need to produce adequate expla¬ nations of all these problems. It seems to me probable that the Doctrine of Addai, the Acts of Sharbel and of Barsamya, the Acts of Guria and Shmona and those of Habbib, form one series designed to celebrate the ancient glories of the Edessene i) ASD 50. 28 INTRODUCTION. Church. The editor of the series worked over the docu¬ ments he used, but he did not invent them. What he may have had for the earlier periods we cannot tell. For the period of the Great Persecution he had the contemporary accounts drawn up by Theophilus, which he incorporated whole, merely adding chronological notices (mostly inaccurate) here and there. I venture to think we are driven to frame some such hypothesis to account for the touches which still make the figures of the Confessors of Edessa live before our eyes. ON THE DATE OF THE MARTYRDOMS. The fact about the dates which is best attested is that Shmona and Guria were executed on Nov. 15 and Habbib burnt on Sep. 2. This is attested by the ancient Syriac Kalendar as well as by all authorities for the Acts. Further, the Edessene Chronicle tells us that in the year 312 — 3 A.D. Konna Bishop of Edessa began to build the Church, which was finished by his successor. It is impossible not to connect this with the Edict of Milan tolerating Christian worship, so that any date for the martyrdoms later than 31 1 is highly improbable. As a matter of fact, Eusebius records martyrdoms in Palestine only between the years 303 and 310 !). Now assuming the good faith of Theophilus, and also (for the reasons given above, p. 17, note) that the clause “year 618 of the Greeks” in N. & G. 69 is an interpo¬ lation, we have the further information that the fifth day after the execution of Shmona and Guria (Nov. 20) was a Sunday 1 2). This fact is also attested both by the 1) A very good discussion of the Chronology of the Martyrs of Palestine is given by H. J. Lawlor in Hermathena XV (1908), pp. 177 — 201. 2) It is perhaps not altogether fanciful to suggest that Theophilus may have had leisure to write on that day, not because it was a Sunday, but because Nov. 20 was observed as the Emperor Maximin’s birthday. See Lawlor , p. 187. 30 INTRODUCTION. Syriac and by d1 in 5. & G. § 41. In other words the Sunday Letter for the year was B, or if a Leap Year CB. This gives as a series of possible years A.D. 292 298 309 315 320. The first is too early, and the last two too late, so we are left with 298 and 309. Sep. 2 is said in Habbib 39 to be a Friday in the Syriac, but a Saturday according to Cl1. Cl2 always leaves out the day of the week. According to the Syriac, therefore, the Sunday letter is B (or CB), according to d1 it is A (or BA). If our Syriac text be accurate the possible years are the same as before, i.e. 298 and 309; and, as Habbib suffered later, it compels us to date Shmona and Guria 298, Habbib 309. If d1 be accurate the possible years are 299 304 310. On general grounds it is likely that the attempted enforcement of pagan sacrifice took place in Edessa later than in Palestine : hence we should ac¬ cept the ‘Saturday’ of d1 in Habbib 39 rather than the ‘Friday’ of the Syriac, and obtain as a final result Shmona and Guria executed . . . Tuesday, Nov. 15, 309 A.D. = A.Gr. 621. Habbib burnt . Saturday, Sep. 2, 310 A.D. = A.G. 621 !). The above examination assumes that the definite and formal dates in our documents are the most liable to error, whether on the part of later scribes or through inaccuracy in the original writer. The year 310 best fits the general situation depicted in the peroration of Habbib (according to the Syriac form), where it is declared to 1) Or 622 if the year begin in Septe?nber\ see below. date of the martyrdoms. 31 be known that Constantine has set out from the extreme West, but the result is still doubtful. This points to some time before Constantine’s victory over Maxeiitins (312): it is noteworthy that no appeal is made by the Martyrs to any previous Edict of Toleration. It is possible that neither the catalogue of Bishops in A. (S’ G. 68 nor the catalogue of Martyrs in § 5 belongs to the original narrative of Theophilus. So far as dates go, the list in § 5 would support the date 309 A.D. for Shmona and Guria, while that in § 68 would be at least neutral, for though Gaius of Rome died in 296, Tyrannus of Antioch did not begin till 303 or later. It should however be pointed out that the list in § 68 is not, properly speaking, a list of synchronisms at all. In the whole of § 67 and § 68 it is not too clear what persons are meant by They’, whether Shmona and Guria, or the Diocletian Martyrs generally, or the Christians of Edessa. And the last sentence in § 68 (“Now some of these things”, &c.) implies that the Four Bishops are men¬ tioned along with Konna of Edessa not because Shmona and Guria suffered in their days, but because other Dio¬ cletian Martyrs suffered iii their days in their dioceses. That Theophilus writing down a really contemporary account of the Martyrs could name the reigning Emperors incorrectly is of course impossible. Very likely he did not name them at all, or the judge, but spoke throughout of “the Emperor(s)” and “the Governor”. In Habbib 39 he may very well have written that the report had come to Edessa that Constantine in the far West had set out to fight ‘with’ Licinius (the present Emperor of the East) against the persecutors. Certainly if he did write in this fashion, it would easily be misunderstood by a 32 introduction. later age, which thought of the Great Persecution as we do, as Diocletian’s, and which came to remember Lici- nius chiefly as one who cashiered Christian soldiers and bullied Christian bishops, and also as the adversary of Constantine the Christian Emperor. In any case the mention of Licinius here probably caused the wrong insertion of ‘Licinius’ in Habbib 3 — 5. The name of Maximinus does not occur anywhere in the transmitted Syriac text, for in Shmona and Guria § 1 the name is written 000.1^000.^3, and this is also the name given to the Governor of Edessa (Mysianus , M v(tizvo(t (f>1 2). Possibly this name is also an insertion of the later editor. When we come to the opening paragraphs it is im¬ possible not to connect the mention of Roman Consuls in the Edessene document with the mention of Consuls at the beginning of Barsamya. The same hand no doubt inserted both sets. Now Barsamya is dated 5 September, 416 A. Gr. (= 105 A.D.), “in the consulship of Commodus and Cyrillus”. This is near enough to Commodus and Cerealis, the real Consuls for 106 A. D., to shew that the editor was using a List of Consuls, not inventing them out of his head *). It is a natural conjecture that this List itself, like the Roman Ravennate Chronicle of 533, was provided with historical notices, e.g. to A.D. 303 Diocletiano VII et Maximiano V is added “His consulib. aecclesiae demolite sunt et libri dominici com* busti sunt”: some notice such as this may have caused the editor of 5. & G. to date his work by the consuls 1) He has further got the years of Trajan wrong; so also with Dio¬ cletian, S. G. 1. DATE OF THE MARTYRDOMS. 33 for 304. I imagine that Theophilus merely gave a year A.Gr. and the Strategi of Edessa as a date. The dates now found in the Syriac at the beginning of the two Acta are A.Gr. 618 and A.Gr. 620. If we might assume that at this period the year began with Ilul = September, instead of ending with it *), this date would be in each case two years too early. Is it possible that the current reckoning of the Seleucid Era in Edessa was two years different from what was afterwards rec¬ koned ? In any case, assuming that the whole narrative covers as short a period as possible, the date at the beginning of S'. & G. should be the year previous to the actual martyrdoms. This, it may be remarked, is an additional reason for regarding the mention of the year 618 in S. & G. 69 as spurious. The drastic treatment of the dates in the transmitted text of Shmona and Guria and of Habbib , here advo¬ cated, may appear somewhat arbitrary. It might seem a simpler solution to reject the tales altogether, as Noldeke and v. Dobschiitz do. I can only repeat that the main reason for continuing to regard the nucleus of the narrative of “Theophilus” as genuine history is its internal character. The simplicity of the tale of the actual executions, and the tone of personal feeling that pervades the perorations, are themselves facts that chal¬ lenge explanation. The perorations did not appeal to 1) This follows from Habbib i8<2, where the K'.’V^rV of the Emperor is evidently Sep. 1. The Syriac word used seems to be identical with that used by the Jews (QTpfcO in c Aboda Zara I 1 for the heathen New Year. In the Edessene Chronicle LXIX (death of Simeon Stylites) the year begins with September; in LXXXI (siege of Edessa) the year ends with September. 5 34 INTRODUCTION. the taste of the various translators. Greek and Arme¬ nian, but they place us in a time when the memory of the Persecution and the terror it inspired was still fresh. If they be not a genuine reflexion of the writer’s age and feelings they are artistic work of very high excel¬ lence, and artistic work of very high excellence is not the characteristic of the literature of Martyrdom. THE PLACE OF THE MARTYRDOMS. The Plan of ancient Edessa generally used by modern writers is that to be found at the end of Wright’s edi¬ tion of Joshua Stylites. It is founded upon C. Niebuhr’s Plan of modern Urfa (1780), supplemented and corrected by later scholars. A proper trigonometrical survey of the town and neighbourhood is much needed, though the preparation of it would be beset with difficulties both from Turkish officialdom and from the fanaticism of the present Moslem population. The main authorities for the compilation of this Note, besides our Martyrdoms themselves, are the Chronicle of “Joshua Stylites” (507 A.D.), the Edessene Chronicle (before 600), Rahmani’s Chronicle (about 1200), and Bar Hebraeus (1285); the accounts of Niebuhr, Badger, and Sachau ; and a set of photographs, some of which are reproduced in my book called Early Eastern Christianity , while that of the Church of S. Theodore (now belonging to the Armenians and called Der Serkis ) is given here as a Frontispiece. The Chronicle of Bar Hebraeus is valuable for the tale of the storming of Edessa by the Atabek in 1 145, of which it incorporates a contemporary account. The 1) Chronicon civile et ecclesiasticum anonymi auctoris .... edidit Ephraem Rahmani (1904). 36 INTRODUCTION. Patriarch Rahmani’s Chronicle contains a detailed list of the Churches and Convents in Edessa [foil. 92 — 94), with some welcome topographical indications : the Chro¬ nicler evidently knew Edessa well, but of course he writes at too late a period to be an unquestionable authority for the age of Diocletian and of Rabbula '). The Edessene Chronicle is partly dependent on Joshua Stylites. According to the unanimous testimony of all our sources the outline of the walls has remained unaltered from very early times. Rahmani’s Chronicler says that “very little” was to be seen of the actual work of Seleucus Nicator (fol. 48), but the ground plan tradi¬ tionally referred to him has remained. It is probable that the designer of the walls planned the main Gates also, but unluckily their names have varied from time to time. Niebuhr marks four Gates, viz. Haran Kapusi, on the S. side, near the SE. corner, Begk Kapusi, on the E. side, where the river leaves the wall, Gengi Kapusi, at the NE. corner, Samsat Kapusi, on the NW. side. Sachau ( Reise , p. 193) enumerates six Gates, Bab Sam- sat (NW.), Bab Esserai (N.), Jeni Kapu (NE.) 1 2), Bek Kapusu (E.), Bab Essakib (W.), Bab Harran (SE.). Badger (1 322) also mentions the Serai Gate, so that it seems to have been made between 1780 and 1852. It does not appear when the New Gate ( Yeni Kapu , ^*4 cCNj) was made. 1) For instance on fol. 58 he assigns the diversion of the Daisan to King Abgar and Addai the Apostle ! 2) See the correction at the end of Sachau’s volume. PLACE OF THE MARTYRDOMS. 3 7 No modern observer speaks of any other Gate on the S. side than the Harran Gate, and indeed the configu¬ ration of the ground makes that tile only suitable point for a town Gate in the S. wall. None of the ancient authorities call any of the Gates by the names of places to which they lead. Rahmani’s Chronicler speaks of the East Gate and of the North Gate. Bar Hebraeus mentions the Sundial Gate p obviously on the S. side. The Edessene Chro¬ nicle (lxviii) mentions the Gate of Beth Shemesh. Joshua Stylites mentions the “Arches” Gate ( Kid'Ll the Theatre Gate and the Great Gate. Of these, the Arches Gate was certainly on the West, for it is called the West Gate in Habbib 35, and the Theatre Gate was opposite it, therefore on the East. It further appears from Joshua § 27 that the river Daisan formerly entered the city near the Arches Gate and left it near the Theatre Gate. We may therefore safely place the latter about half-way between the Harran Gate and the Bey’s Gate. I venture further to identify the Great Gate of Joshua not with the Harran Gate at the SE. corner, but with the Bey’s Gate [Beg Kapusi) on the E. side. For Joshua § 36 says . P^=3 i 0,1=3 <^OCD This must mean : “There was a breach in the wall at a place south of the Great Gate” '). Now nothing in Edessa is south of the Harran Gate, so that Joshua’s Great Gate must be further to the north. It will be seen presently that other indications confirm this. 1) For the idiom, see Noldcke § 247. 38 INTRODUCTION. There is every reason to suppose that Bar Hebraeus’s Sundial Gate is the Harran Gate, and I believe also that this is the Gate called Gate of Beth Shemesh in the Edessene Chronicle (LXVIIl). Is it not possible that a Sundial may have been called Sun House as well as Hours’ House? The name only occurs in this one passage. If we make these identifications, and only if we do make them, the topographical notes in Rahmani’s Chro¬ nicle become comprehensible and the account of the disposition of the Persian Army in Joshua’s chapter about the siege of Edessa becomes clear and consistent. With the knowledge thus gained and confirmed we shall be able with some confidence to locate the places where Shmona and Guria and Habbib were martyred, and where the Shrine stood in which their Coffin was pre¬ served. The List in Rahmani’s Chronicle names the following Churches in and around the city *). 1. Church of Thomas the Apostle in the SW. corner of the city [= the Great Church]. 2. Ch. of the XII Apostles in the E. quarter. [By the “Great Gate”, says Josh. Styl. § 43]. 3. Ch. of Sergius the Martyr by the E. Gate. [The Churches of S. Sergius and of S. Simeon were out¬ side the walls, Josh. § 31]. 4. Ch. of John the Baptist in the W. quarter, with red marble pillars. 5. Ch. of Stephen, in the middle of the city, which had been a Synagogue. [This, as all travellers agree, 1) Additional notes or identifications not contained in Rahmani’s document are added in square brackets. I abbreviate the wording of the Chronicle. PLACE OF THE MARTYRDOMS. 39 corresponds to the Ulu Jami, the Great Mosque in the centre of the town, with its ancient octa¬ gonal tower]. 6. Ch. of Sergius in the E. quarter a little south of the Apostles. 7. Two Churches of Theodore, one (a) in the Citadel (Kliflaw), the other (b) outside, in the hill, in the W. quarter, on the Serug road. [N°. jb is doubt¬ less the building shewn in the Frontispiece, now called also Der Serkis and Khidr Elias]. 8. Two Churches of the Confessors, one (a) in Ramath Dauke, with monastery attached, the other (b) for Guria, Shmona, and Habbib, by the N. Gate of the city. [N°. 8 a is the Church mentioned by Bar Hebraeus, n°. 8b is our Shrine]. 9. Ch. of Cyriacus, NE. of the Confessors. 10. Ch. of James of the times of Julian in the village of [= “Garamoosh” [Badger I 325), the Armenian village three hours from Urfa]. 11. Churches of SS. Cosmas and Damian: [a) Cosmas in the S. quarter, outside in the plain; (b) Damian above, at the top of the bare cliff near the city. The Saints themselves are buried in these Churches. 12. In the hill many monasteries, inch B.V.M., the “Orientals”, and two of S. Barbara. 13. Monastery of Kubbe, at the foot of the hill, south of the Ch. of Cosmas. 14. Monastery of the Exedra, on the top of the hill, and others. 15. Ch. of James in the midst of the hills, where there had been a great altar of the pagans which is standing to this day in the monastery 40 INTRODUCTION. itself. [This is evidently the Der Jackub (Surp Hagop), where Sachau found the ancient inscription of a,(j.K f°r “they had brought in the bones of all the martyrs which were around the city” (Josh. § 59), to save them from injury. Indeed the other Shrine of the Confessors in Rdmath D alike (Watchmen’s Mound), with its monastery attached, seems to have been designed as a place of refuge for the relics and the monks who dwelt outside in times of peace. In any case, so far as our Acts of Martyrdom and the story of Euphemia is concerned, the Shrine of the Confessors evidently means the Shrine near the N. Gate. In only remains to locate Beth Alah Kikla itself. From what has been said the area of possibility has been greatly narrowed. The Martyrs go out alive through the 1) Wright’s Josh. Sty l p. 51. K&lCUxJL is “bad roads” rather than “steep descent”: probably the name denoted the ford over the Germish Chai. 6 42 INTRODUCTION. W. Gate, and their bones are preserved near the N. Gate : Beth Alah Kikla must be somewhere NW. of Edessa. On general grounds, as the executioners took out Shmona and Guria, and afterwards Habbib, by the West Gate, and appear to have returned the same way, we must look for the places of execution nearer to this Gate than to the Gate further north, from which their Shrine afterwards was approached. The very mention of the Coffin (S. & G. 6y) leaves room for the subsequent removal of the hallowed bones. The data given by the Acts are 1. Shmona and Guria were beheaded at Beth Alah Kikla, a rising ground N. of Edessa (r?i but SW. of the spring that flows into the town (5. & G. 56). 2. Habbib was burnt by the Cemetery of cAbshlama bar Abgar (H. 35). 3. Habbib’s body was carried in procession by the Christians to the grave of Shmona and Guria in Beth Alah Kikla (H. 38). 4. The Shrine of the Confessors was near a hill [Eup hernia 29, end), and was itself situated on rising ground [Enp hernia 13, 33). It was therefore not just out¬ side the N. Gate. We know nothing otherwise about cAbshlama’s Ceme¬ tery, but as the Testament of Ephraim (§ xi) commands the body of Ephraim to be buried with the poor strangers in common ground, and the reputed grave of S. Ephraim actually is at Der Serkis (i.e. the Church of S. Theodore), it is not unlikely that this Cemetery was in the neigh bourhood of Der Serkis. Further, just before, at the end of § x of the 7 e st anient we read PLACE OF THE MARTYRDOMS. 43 . K'ioiKir) rtA.i \ rtf^AxiA&a r^.U. crA . rtf&An&f) pdl^QQ^Ao *. K'pCa OCT3 K'Ta^A “Leave it (Ephraim’s dead body), cast it on the dunghill, for it cannot be conscious of honour. Wealth is fitting for the wealthy, and for the poor the dunghill”. Is not this reiterated mention of the dunghill a refe¬ rence to Beth Alah Kikla, the “Dung God’s Shrine”? On the other hand Beth Alah Kikla is said to be N. • • of Edessa, yet SW. of a certain spring which goes into the town. This spring is very likely the same as is called Keris Suju (i.e. Canal River) in the corrigenda to Sachau’s volume, as it comes into Urfa from the NW. The lines of the aqueducts shewn on the plan are those by which according to Wright the water came from Telzema and Maudad (Josh. Styl. § 87): this is probably the aqueduct called Dschansach by Sachau’s correspon¬ dent. The spring (rtl^o.n.’sa) mentioned in S. & G. 56 may be what Badger (l 328) calls Sooleiman Pyar, a spring where Addai is said to have baptized the first converts. With all this it must be remembered that when our documents were written the Daisan did not go round Edessa as it now does, but still flowed through the town. Moreover, there is rising ground inside the walls opposite the Samosata Gate and for some distance to the SW. (Sachau, 192, 194), so that it is clear the ancient course of the river must have been south of this again, probably south of S. Stephen’s Church. The general result is to locate the place of Habbib’s martyrdom not very far from Der Serkis, where S. 44 INTRODUCTION. Ephraim’s tomb is, but further to the north, i.e. on the extreme right hand of the Frontispiece to this book. The place where Shmona and Guria were executed must have been yet further to the north, beyond the present course of the Daisan and near Solomon’s spring. It is here that we must look for remains of the Shrine built about A.D. 350 by Bishop Abraham , the Shrine to which Euphemia went up with her mother and the Goth, and to which she returned in so strange and dramatic a fashion. (TO FACE THE PLAN OF EDESSA). Gates of Edessa — Urfa: 1 . Bey’s Gate ( Bek Kapusi ) — Great Gate [Josh. Styl.]. 2. Theatre Gate [Josh. Styl.]. 3. Harran Gate = Sundial Gate [Bar Hebr.] ; G. of Beth Shemesh [Chron.]. 4. W. Gate ( Bab Essakib, Daghlar Kapu ) — G. of the “Arches” [Josh.; Habbib]. 5. Samosata Gate = old N. Gate. 6. Serai Gate. 7. New Gate (Yeni Kapu). The Cross close to N°. 3, marked Conf., is the New Church and Monastery of the Confessors, mentioned by Bar Hebraeus and in Rahmani’s Chronicle. The large Bastion just south of it is probably what was called Ramath Dauhe (i.e. Watchmen’s Height). Near the Church of S. Theodore in the Citadel is marked the position of the two great Columns. S. Theodore outside the walls, now the Armenian Monastery of S. Sergius ( Der Ser/eis), is the building shewn in the Frontispiece. The modern Armenian Cathedral must be very nearly on the site of S. John Baptist. ABOUT EUPHEMIA AND THE GOTH. § i. From the time that I first read the story of Euphemia in the Nitrian MS. I had no doubt that it was of Syriac origin. The Greek version had been known to me chiefly from a fragmentary MS. at Emmanuel College, Cam¬ bridge, but it seemed altogether inferior and secondary compared with the Syriac. As however Professor von Dobschiitz, in his admirable edition of the Greek text of the tale, maintains that it, and not the Syriac, is the original, it is necessary to indicate the reasons which incline me to the contrary opinion. A tale about the local Saints of Edessa might natu¬ rally be supposed to have originated there, and any - other hypothesis needs either express external testimony or strong internal arguments. It will be convenient to begin by enumerating Prof. v. Dobschiitz’s arguments, to be found on p. L of his edition. “i. Not only is no Syriac text of Euphemia known, but also there is no reference to the tale in Jacob of Serug’s Homilies or in the rest of Syriac literature”. The production of two Syriac MSS., one of them as old as the 9th century, is not perhaps a complete answer to this, though it goes a considerable way. It may be ABOUT EUPHEMIA AND THE GOTH. 49 further urged, with reference to the silence of Jacob of Serug and others, that the same argument might be used to prove that the “Hymn of the Soul” in the Acts of Thomas was not a Syriac original. For Jacob com¬ posed a Homily on the Palace which S. Thomas built, but he is silent about the great Hymn; and, on the other hand, the Hymn is both extant in Greek and is also the subject of an Encomium by Niceta of Thessa- lonica ! The fact is, that we do not possess an approxi¬ mately complete or representative Corpus of Syriac Literature, in the same way that we possess it for the Byzantine Greek and Mediaeval Latin Literatures. What we have is the contents of a single monastic Library, supplemented by late collections from elsewhere. And the special interest of the story of Euphemia is its secular tone ; it is only in some of the secondary Greek texts that the heroine becomes a nun. “2. All the Greek texts go back to a single literary work, which shews no trace of being a translation from the Syriac”. Now that the Syriac is before us we see that the “translation” was so free that linguistic traces of a foreign style are scarcely to be looked for. The Syriac is as much fresher, less ecclesiastical, more like a folk¬ tale, than v. Dobschiitz’s text, as that text is than those in the Menaea. “3. The quotations from the Bible are taken from the Septuagint”. In the Syriac the few Biblical passages that occur (mostly from the Psalms) are given in according with the Peshitta. The translator, whether Greek or Syrian, simply made use of the familiar words of his own Bible. 7 50 INTRODUCTION. “4. The mother and daughter, Sophia and Euphemia, have Greek names”. But these names are hardly more foreign on Meso¬ potamian ground than “Sophia” and “Euphemia” are in England. Euphemia was the name of a famous martyr, actually mentioned in the Edessene Chronicle (§ 83) 1). “5. The character of the Tale itself is that of the Greek Novel”. Our knowledge of the range of Syriac Literature is less than that of Greek Literature. Yet Syriac Literature includes Romances such as the Acts of Thomas, with its brilliant sketches of the minor characters. The dull “Romance of Julian” is a regular historical novel. But indeed the East has always been famous for the telling of Tales: the Story of Euphemia is at least as near to the Arabian Nights as it is to the tale of Clytemnestra and Cassandra. It is, I venture to think, obvious that no safe con¬ clusion can be reached from these general considerations. The nature and provenance of the Syriac MSS are equally indecisive. The 9th century Nitrian MS was written by one Simeon, who copied it for his own use (Wright, CBM 1110): the contents of the volume are miscellaneous, same of the stories being taken from such well-known sources as the Historia Lausiaca, no doubt at second or third hand. Most of the twenty-six items are found elsewhere in Syriac MSS, and one of them, the History of Simeon of Kephar cAbdin, is certainly of Syriac origin 2). The Paris MS was written in the 1) Curiously enough, here also is a mention of a miracle wrought in connexion with a Coffin (gluskma ). 2) See also Appendix I for another Syriac tale from the same MS. ABOUT EUPHEMIA AND THE GOTH. 51 13th century at Antioch, and it also is a collection of edifying tales. At the end of Euphemia we find a Note, as follows : “Here endeth the story of the miracle that the Holy Confessors performed on Euphemia who was betrothed to the Goth in Edessa: it was copied from an old book that was written in the royal city of Constantinople by John the Monk and Recluse”. So far as it goes, this certainly would suggest a Greek origin for the tale, but the Note is not confirmed by the Nitrian MS, written some three or four centuries earlier. And even if it be regarded as a serious piece of information, it does not legitimate the existing Greek text, as edited by v. Dobschtitz, against the Syriac text here published J). Thus everything goes back to the internal evidence of the two texts themselves. And here I feel the verdict is clear and unhesitating. The Syriac text is in every way more original and superior to the Greek. It is difficult indeed to offer any formal demonstration, for neither text is a literal rendering of the other. But to give an idea of the characteristics of the two texts I will set down a few quite literal English translations of parallel passages side by side. 1. (The actual beginning of the tale). Syriac . Greek. § 4= v. Dobschiltz , p. 150. In the year 707 by the In the year 707 of Alex- 1) The phrase “royal city” ( Medinath Malkutha ) makes the conjecture possible that this part of the colophon originally belonged not to Euphe- tnia , but to the Story of the Wife of Patricius and the Merchant of Harran (Appendix I). 52 INTRODUCTION. reckoning of the Greeks the [Huns] *) had come forth [into the Roman territories], and they captured many captives and laid waste the country and came as far as Edessa. And Addai the Stratelates of those days did not give permission for the Foederati to go out against them, be¬ cause of treachery in the midst, and for this cause the armies of the Romans came down and lived in Edessa some time. Now a certain Goth ander King of Macedon that evil and destructive nation the Huns, having come into the Roman territories and laid waste various places and taken many captives, arrived as far as the district round Edessa. \no equivalent ]. For this cause armed for¬ ces gathered from various regions occupied the city and lived in it a long time. Now a certain Goth [The point to notice is that the clause which gives definite local and political information is absent from the Greek]. 2. (Sophia’s ejaculation, when against her better jud¬ gement she consents to let § 10 = “God of the orphans and the widows, come to my help ! My God, this business the Goth marry Euphemia). v. Dobschiitz , p. 1 5 67 . “Thou, Lord, Father of the orphans and Judge of the widows, look upon what 1) The Nitrian MS has ‘Persians’ on an erasure. The Huns are men¬ tioned below in § 35, along with the Persians, as in the Greek. On the Syriac text of the first sentence see the Note on the passage. ABOUT EUPHEMIA AND THE GOTH. 53 is entrusted to thee, to is being done, and do not whom alone it is revealed !” look away from my orphan child or helpless me! For trusting in thy good Pro¬ vidence I am enduring the marriage of my unhappy daughter with an unknown man, who calls thee to be witness and surety of his own promites”. [On p. LIV Prof. v. Dobschiitz says: “Besondere Freude aber hat der Verfasser offenbar an der Komposition von Gebeten .... hier fiihrt ein griechischer Theologe die Feder”. Now that we have the Syriac original, we can see exactly how much is due to the Greek theologian !] 3. (Sophia visits the shrine of the Confessors with Euphemia and the Goth before they go away to his home). § 13 = And when they had sealed their prayer with a tear, the mother of the girl drew near and took hold of her by her right hand and set her upon the Coffin ’) of the Con¬ fessors, saying to her false v. Dobschiitz, p. 1 5814. And as they were stand¬ ing by the reliquary of the Holy ones she said to that wicked man : “Not otherwise will I entrust my daughter to thee , except first having 1) The Syriac has Gluskma , an adaptation of yAu