¹ Clarke, G. W., ‘Syriac inscriptions from the Middle Euphrates’, Abr Nahrein, XXIII, 1984–85, 73–82.Google ScholarThe author wishes to thank Professor Clarke for sending her the photographs of the inscriptions and other information pertaining thereto, the Rev. Professor John Emerton for the assistance which he has given her in the preparation of this paper, and Dr. Andrew, Palmer for preliminary suggestions.Google Scholar

² Muraoka, T., ‘Two Syriac inscriptions from the Middle Euphrates’, Abr Nahrein, XXIII, 1984–1985, 83–9.Google Scholar

³ Montgomery, J. A., Aramaic incantation texts from Nippur (Philadelphia, 1913), 27–8,CrossRefGoogle Scholardiscusses the shortcomings of using palaeography as a dating method and emphasizes the role of scribal idiosyncrasy in the execution of scripts.Google Scholar

⁴ Clarke, , op. cit., 79, cites the usage of crosses over the entrance of the tomb-chamber and over the actual tomb itself to support the proposed Christian burial. See plate I (a).Google Scholar

⁵ This measurement was given in private correspondence between Clarke, and Hunter, (24 September 1986), although the actual distance between the two inscriptions does not seem to have been recorded.Google Scholar

⁶ Hatch, W. H. P., An album of dated Syriac manuscripts (Boston, 1946), 11,Google Scholarnotes that the Syrians made an ink out of lampblack mixed with oil or gum.Google Scholar

⁷ The extent of the weathering is difficult to determine. In private correspondence (14 June 1984) Clarke suggested to Hunter that the tomb-chamber had been recently opened by villagers, resulting in some deterioration.Google Scholar

⁸ See plate 1(b).

⁹ In the absence of measurements being recorded in situ, the tracings of the inscriptions were placed over a grid and the results were recorded.

¹⁰ Clarke in private correspondence with Hunter (24 September 1986) thinks that a local lad tried to write his name AHM.. in ‘English’script or may have made an attempt to commence the alphabet.Google Scholar

¹¹ Muraoka, , op. cit., 88, is convinced that a line exists, thus making Inscription II nine lines in total. We fail, however, to detect any traces of an intermediate line, so that the sum stands at eight.Google Scholar

¹² An exact estimate cannot be made, due to the deteriorated state of Inscription II, where no single complete line remains.Google Scholar

¹³ See plates II and IV (a).

¹⁴ The terms majuscule and minuscule first appear to have been used by Land, J. P. N., Symbolae Syriacae (Lugdano-Batavae, 1862)and were adopted by Enno Littmann, Syria: Publications of the Princeton University Archaeological Expeditions to Syria in 1904–1905 and 1909. Division IV. Semitic Inscriptions (Leiden, 1934), p. x. He notes that Land's majuscule script was about the same as Estrangela and his minuscule script was similar to Serta. Hence we have elected to use the terms majuscule and minuscule respectively to indicate a script with formal characteristics, showing an affinity with Estrangela and a script with cursive tendencies suggestive of Serta. Whilst Land coined the terms in respect to manuscript material, Littmann's publication concerned incised inscriptions, and therefore would be somewhat appropriate to Inscriptions I and II.Google Scholar

¹⁵ Muraoka, , op. cit., 83, reads ṦL..., thereby interpreting the opening stroke as that of Shin and the Seyame as the upper points of an open, unfilled Shin; an example of which is found in 1. 2. If the letter were read as such, the spacing between the dots would suggest that it was cramped. The interpretation of the two dots as Seyame is, of course, supported by the preceding word KLHWN which would be accompanied by a plural, masculine substantive.Google Scholar

¹⁶ Hatch, , op. cit., plate CXXII, provides several examples of a Waw introduced by an oblique stroke that turns into a short, vertical stroke connected to the base line to form a closed character. Furthermore, in contrast to the common practice, Waw appears to be ligatured on several occasions to the consecutive Taw.Google Scholar

¹⁷ Muraoka, , op. cit., 83, reads YRW..., but the angularity of the fourth letter would argue against it being a Waw, as the letters in KLHWN and on 1. 2 WBŠMH indicate a cursive trend.Google Scholar

¹⁸ ibid., 84, suggests that the top horizontal bar of the Shin should be restored, on account of the letter's form in 1. 5. Both the open and closed forms of Shin do occur in inscriptions as seen in Littmann, op. cit., Inscription 42 from Der Sim'an and Inscription 52 from Kefr Nabu, both of which date from the sixth century A.D.

¹⁹ ibid., 83, sees two letters between the Resh and the final Alaph, hence producing the reading MRYM’. On the grounds of space, it would seem difficult to accommodate a penultimate Mim, given the size of the two examples which occur in 1. 2.

²⁰ idem reads ’.ZQYP’, but on p. 85 he states that the ‘excessive space between the initial Alef and the Qof... would speak against such an interpretation.’Furthermore, Zain always seems to be written as a short, vertical stroke and it is difficult to visualize the elongated, horizontal stroke leading into the open curved form as this letter.Google Scholar

²¹ A survey of manuscripts in Hatch, , op. cit., and inscriptions in Littmann, , op. cit. and Pognon, H., Inscriptions semitiques de la Syrie, de la Mésopotamie et de la region de Mossoul (Paris, 1907)Google Scholar, does not indicate that the right-hand loop of the Semkath was written larger than the left- hand loop, as would be the case here.

²² The difficulty with this interpretation is that the above survey also establishes that the connecting stroke joins the letter Semkath at its base line and not at the top of the right-hand loop.

²³ Muraoka, , op. cit., 83, reads.. YN, by which he has interpreted the initial Alaph as two letters, albeit undeciphered. Although the penultimate letter could be a Yodh, the clear example of a Yodh ligatured to the following Nun in 1. 5, acts contrary to this reading.Google Scholar

²⁴ idem similarly reads M.. and comments on p. 85, ‘we first thought we could see M‘DR “helper”, but so little is left visible of the latter half.’ Muraoka may, therefore, lend cautious support to our suggestion.

²⁵ idem interprets these two letters as N’ concluding the word M...N’. Whilst our reading does differ, agreement has been reached on the final Alaph. I have not been able to decipher the first letter which Muraoka has read as a Mim. Nor would I specify how many letters occur before the penultimate character; on the grounds of spacing, three or four might be written. Nor would I specify what word division might occur as several options could apply. M‘(D)R(NY’) L’could be proposed or one could even suggest M‘(D)R (S‘D’) L’. Both readings would be satisfactory in a contextual sense.

²⁶ ibid., 85, is’slightly concerned about the gap between the end of the base line of the Beth and the Dalath’in ’BDYN. There does not seem to be any suggestion of ’BYDYN which Muraoka quite promptly rejects and the slightly irregular spacing may be attributed to scribal idiosyncrasy.

²⁷ Littmann, , op. cit., 41.Google Scholar

²⁸ Hatch, , op. cit., 24, discusses the role of the medium in the production of scripts and in doing so distinguishes between epigraphy and palaeography. In adhering to his distinction, the latter term has been used in connexion with Inscriptions I and II since they have been painted or written, rather than incised, on the rock medium.Google Scholar

²⁹ Muraoka, , op. cit., 84, comments that the letter Mim, which is closed, shows interesting variations on the upper oblique stroke—ranging from the almost horizontal example to those slanting in different degrees. As such, he proposes that the letter represents an intermediate stage between classic Estrangela and classic Serta. Whilst I agree with this finding, I would also point out the similarity of the letter to several specimens in inscriptions from the thirteenth century A.D.Google ScholarSee Littmann, , op. cit., Inscription 28 from Dēr Sim’ān and Pognon, H., op. cit., Inscription 94 (plate XXXVIII) from the Convent of St. Abai near Mardin. The form of the letter where the upper oblique stroke proceeds from the top left to the top right instead of descending to the base line, is also found in a Melkite manuscript from the thirteenth century A.D. See Hatch, op. cit., plate CLXXIII.Google Scholar

³⁰ ibid., 83, claims that the letter Resh in the inscription has a’shape of a reversed C’. We would describe the forms of both the letters Dalath and Resh as being intermediate between the Serta and the angular shape of Estrangela. See Pognon, op. cit., Inscription 94 (plate XXXVIII).

³¹ ibid., 84, notes that the He departs from the classical Estrangela form and approaches the classic Nestorian style of two left-hand downward strokes joined to form a circle. The closed form of He, with a straight connecting stroke and stem is found frequently in inscriptions; appearing as early as A.D. 491 at Basufan. See Littmann, op. cit., 39. Apart from occurring in Estrangela and Nestorian manuscripts, the He can be seen in a Melkite manuscript from A.D. 1045. See Hatch, op. cit., plate CLXXXIV.

³² The Alaph at the end of 1.2 is the only definite example of the majuscule type, although the first letter of the second word in 1.3 could be also. The semi-majuscule type is represented by the initial Alaph in 1. 3 and the first Alaph in 1.4 and its cursive trend would appear to uphold the transitional characteristic of the inscription's palaeography. The two types of Alaph are found in the thirteenth century inscription from the Convent of St. Abai, near Mardin. See Pognon, op. cit., Inscription 94. The semi-majuscule Alaph first appears in a Nestorian MS from A.D. 719–20, but it also frequently occurs in the Estrangela script; with the earliest example dating from A.D. 740. See Hatch, op. cit., plates CLXII and LXXIV and also plate CLXXXIV, being a Melkite manuscript from A.D. 1045 wherein the semi-majuscule and minuscule types are written.Google Scholar

³³ Like, Muraoka, op. cit., 85, we would group together the two Alaphs at the end of 11.4 and 5 respectively, noting that they are the product of a single stroke, rather than two. We have termed these characters minuscule and they appear to belong to the Serta tradition, and together with the majuscule and semi-majuscule types appear in a Serta manuscript dated from A.D. 1056. See Hatch, op. cit., plate CXXII. Finally, three forms of Alaph are also found in a graffito from Dēr Sim’ān in northern Syria. See Littmann, , op. cit., Inscription 32.Google Scholar

³⁴ Littmann, , op. cit., 37, whilst commenting on the mixture of minuscule and majuscule letters in an inscription from Der Sim'an says that the repetition of this phenomenon, ‘shows that the men who carved inscriptions or graffiti endeavoured to use the majuscule script because of its monumental character, but were influenced by the minuscule script which was then in use for manuscripts.’Google Scholar

³⁵ Littmann, , Semitic inscriptions (New York, 1904), 33, says that may be considered as an intermediate stage between and the later form of Shin.Google Scholar

³⁶ Hatch, op. cit., Plates CXCVIII and CC, being Palestinian Melkite manuscripts written in A.D. 1030 and 1118 respectively. The examples cited by Muraoka, op. cit., 84, do exhibit a triangular shape, but otherwise bear little resemblance to the discussed Shin form since they are filled in.

³⁷ Gwynn, J., The Apocalypse of St. John in a Syriac version hitherto unknown (Dublin, 1897), p. cxii,Google Scholar discusses the features of the script of later Estrangela. He particularly notes the closed forms of He, Waw and Mim as well as the curved, rather than rectangular shapes of Dalath and Resh.

³⁸ Hatch, op. cit., 30.

³⁹ ibid., 28–9, outlines the distinguishing features of the Melkite script and provides notes on the individual characters in pp. 31–7.

⁴⁰ Segal, J. B.,‘Four Syriac inscriptions’, Bulletin of the School of African and Oriental Studies, xxx, 1967, 302–4,Google Scholar discusses a grave inscription from Şuayip-şehri, near Edessa, and infers that it was erected by members of the Melkite church.

⁴¹ Muraoka, op. cit., 85.

⁴² Smith, R. Payne, Thesaurus Syriacus (London, 1879–1901), vol. II, col. 1632.Google Scholar

⁴³ ibid., col. 1609.

⁴⁴ ibid., col. 1582.

⁴⁵ Littmann, op. cit., 21, in a discussion of the spelling of GYWRGYS by GYWRG’. Similarly, we find ‘Stephanos’spelt in a variety of ways including ‘STPN’. See Littmann, op. cit., 52.

⁴⁶ Caquot, A., ‘Nouvelles inscriptions Araméenes de Hatra’, Syria 29, 1952, 98, in discussion of the name HWBS’.Google Scholar

⁴⁷ Prentice, W., ‘Magical formulae on Syrian lintels’, American Journal of Archaeology, x, 1906, 145,Google Scholarnotes the frequency of βοθει or βοθησον on lintels in Syria. He also mentions the invocation of saints in formulae such as, ’Aγιε Σεργι (Proper Name) + βοθησον. E. Littmann, op. cit., 1904, 28, discusses an inscription from Bāḳirha dated A.D. 546 which invokes for the salvation, L‘WDRN’, of the monks’souls. On p. 45 in an inscription from Mektebeh in A.D. 508–9 one reads, ‘LH’L KX ’DRN BR BṢṢ ‘O God, (mighty) above all (?), helper of Bar Bassos’. On p. 46 Littmann claims that ‘DRN is the Syriac equivalent of the Greek. H. Gollancz, The Book of Protection, being a collection of Syriac charms (1912, reprint London, 1967), makes numerous references to divine assistance in the individual amulets which have been translated.Google Scholar

⁴⁸ Noldeke, T., Compendious Syriac grammar 1904, reprint Tel-Aviv, 1970, p. 211.Google Scholar

⁴⁹ See plate III, IV(6).

⁵⁰ Muraoka, op. cit., 86, reads.TT. It is doubtful whether another letter occurs before the first legible character which is aligned with the right-hand, albeit crude, margin of the inscription. The interpretation of the initial character as a Taw probably stems from the faint foot which may be seen. However, in 1.2 Muraoka reads D'S’L where the unattached Alaph also has a foot. Unfortunately, the initial letter of 1. 1 is too blurred to allow a definite assessment, but the reading.TT, being dependent on the presence of “TT midway in the line, is tenuous since the latter is only speculative.

⁵¹ This form of Alaph first appears in a Serta ms from A.D. 731, but is frequently attested after this date. It also occurs in both the Melkite and Nestorian scripts, being found in manuscripts from A.D. 1045 and A.D. 1259 respectively. See, Hatch, op. cit., plates XCV, CLXXXIV and CLXXIII.

⁵² Muraoka, op. cit., 87, reads GR... as one word with BR.

⁵³ The majuscule style of Gamal appears in Serta manuscripts as early as A.D. 731 and dominates until the mid twelfth century when a cursive form emerged. See Hatch, op. cit., plates XCV-CXXV; CXXVI-CLIX. This transition does not seem to have taken place in the Nestorian or Melkite scripts. Similarly, inscriptions show the same conservative tendency wherein the majuscule style is preferred.

⁵⁴ Muroaka, op. cit., 86, proposes that two letters precede the initial Alaph.

⁵⁵ The usage of the two types of Alaph occurs in an inscription from Der Sim'an. Littmann, op. cit. (1904), 29, says that‘the initial Alaph is only a perpendicular stroke, whereas the final Alaph has kept its old form’. We would specify that the designated ‘old form’is the semi-majuscule Alaph. Ibid., Inscription 46, indicates that this form did appear in the initial position.

⁵⁶ Hatch, op. cit., 38, notes that, ‘the earlier form of the letter was also employed in the Serta style of writing until the second quarter of the twelfth century and occasionally thereafter.’ This form of final Kaph is distinguished by a serpentine leading into the straight downward stroke.

⁵⁷ This Ṳadhe is unique and one may ask whether its form has resulted from difficulty of writing on the friable limestone medium.

⁵⁸ Muraoka, op. cit., 88, has read a Heth but otherwise does not comment.

⁵⁹ idem discerns a Taw, possibly on the basis of the line, which I have interpreted as an extended, oblique upper stroke of Mim, being the ligature from the letter Heth to the Taw. But this interpretation also ignores the clear base line between the penultimate and final characters.

⁶⁰ This elongated stroke would lend support to the earlier reading of the penultimate letter of LM’as Mim.

⁶¹ If it did form the final letter of Š(L)MT(?), then we could expect to read either a He or an Alaph.

⁶² idem reads the letter as a He, but the initial upward stroke would have to be supplied even if it was of the ‘open’form. On the other hand, the combination of open and closed forms of Waw occurring in the same inscription can be seen in an example, dated to the twelfth to thirteenth century A.D., from the Convent of St. Behnam near Nimrud. See Pognon, op. cit., Inscription 75.

⁶³ Muraoka, op. cit., 88, postulates that the letter could be either a Qoph or a ‘clumsily formed’Waw. He questions the first reading on the grounds of the letter's angularity and it is for this reason that we reject the second suggestion. The contiguity of the closed He with the following letter as seen in a Serta MS dating from A.D. 902, may suggest that the character was a closed He wherein the left- hand loop has largely deteriorated. See Hatch, op. cit., Plate CXIII.

⁶⁴ This interpretation by Muraoka, op.cit., 86, may be accepted on purely palaeographical grounds, whereas his alternate suggestion of a Yodh or a Nun seems to be most unlikely.

⁶⁵ idem reads the initial letter as a Nun, on the basis of the short, upright stroke which we have interpreted as part of the triangular body of the Shin.

⁶⁶ The slant of the Alaph reproduces that of D’ŠL in 1. 3, but it may be noted that the slant of the Alaph in N/YŠL in 1. 7 is leftward, thus showing the inconsistency of the scribe's hand.

⁶⁷ idem reads N'MN, thus interpreting the third letter as a Mim; probably from the small circle which may be seen at the junction between the upright and the base strokes.

⁶⁸ idem reads the Lamadh as an Alaph.

⁶⁹ idem reads the Mim as a Lamadh.

⁷⁰ idem reads the Kaph as a Beth.

⁷¹ idem reads NS..., thus interpreting the initial stroke as a Nun, whereas it is, in my opinion, part of the Šadhe. This form of the letter occurs in Serta manuscripts until the mid twelfth century when there appears to have been a 90 degree shift in the initial stroke to produce the subsequent’3’form. See Hatch, op. cit., Plates CXXIV and CXXVI.

⁷² idem reads DMY.

⁷³ Hatch, op. cit., 46, divides the history of the Serta script into two major periods; the first from A.D. 700 to A.D. 1100 and the second from A.D. 1100 until the sixteenth century. On p. 38, in his comments on the final Kaph, Hatch notes that,‘the earlier form of the letter was employed in the Serta style of writing until the second quarter of the twelfth century and occasionally thereafter. The later form of final appeared in the second quarter of the twelfth century and continued in use down to the end of the sixteenth century.’In the same period the Sadhe appears to have undergone a re- orientation, for the letter was written as rather than as , as occurs in earlier manuscripts. See Hatch, op. cit., plates XCVII, CXXIV and plates CXXVI, CXXXVI.

⁷⁴ ibid., 36, notes that Shin was written in the Estrangela fashion in the oldest Serta manuscripts, being those up to the last quarter of the eleventh century A.D. when the cursive style appeared and superseded the majuscule Shin. The same phenomenon seems to have occurred with the letter Gamal, where a cursive style appeared in the middle of the eleventh century and replaced the Estrangela style Gamal that was previously used in Serta manuscripts. See Hatch, op. cit., plates XCVII-CXXV and plates CXXVI-CXXXVI.

⁷⁵ idem, plates CLXXXIV-CXCVII, although on p. 29 Hatch cautions that Melkite manuscripts are not well represented, with a paucity of dated manuscripts prior to the sixteenth century.

⁷⁶ Payne Smith, op. cit., col. 4143, notes under the entry ŠYR', Shira, ‘nomen coenobii Turabdinensis ad ripas Euphratis’as quoted by Bar Hebraeus. If ŠYR'is a title, then the inscription could commence with a statement of rank before the proper noun. See Littmann, op. cit., Inscriptions 15, 42–4, 62, for examples which commence with the formula: ’I’ + title + name.Google Scholar

⁷⁷ If ŠYR'is interpreted as a title, then BRGR could be read as a single word as Muraoka has done.

⁷⁸ Nöldeke, op. cit., §266, p. 208.

⁷⁹ See the comments by Muraoka, op. cit., 87–8, regarding the idiosyncrasy of this construction.

⁸⁰ See Littmann, op. cit., Inscription 8 and Pognon, op. cit., Inscriptions 95, 102 and 103 for examples of the plural Imperative. By contrast, the Imperfect construction occurs in Littmann, op. cit., Inscription 62 and Pognon, op. cit., Inscriptions 24, 26, 54, 67, 68, 75, 92 and 94. Littmann, op. cit. (1934), 20 comments on variants of the formula.

⁸¹ See Littmann, op. cit., Inscription 62 and Inscriptions 24, 26, 54, 67, 68, 75, 92 and 94 in i Po gnon, op. cit.

⁸² Clarke, op. cit., 79.

⁸³ idem.

⁸⁴ idem. The pieces of pottery and glass are presumed to be contemporary with the early site, although intermittent occupation may have continued even up to modern times. A graphic account of the type of habitation is given by Bell, G. L., Amurath to Amurath (London, 1924), 39.Google Scholar When in the area of Qal'ah Najm she ‘rode from the Euphrates up a bare valley... to a great cave inhabited by Arabs. It contained three chambers, the opening of which had been fenced in by the latest inhabitants with screens made of rushes.’

⁸⁵ Goossens, G., Hierapolis de Syrie: Essai de monographie historique (Louvain, 1943), 195–9Google Scholar discusses the crossing points of the Euphrates and, in particular, Qal'ah Najm. In addition to the overland traffic between Seleucia and Antioch, river trade may also have occurred especially since the situation of Djebel Khaled,‘on the south side of a range of hills through which the Euphrates forces a swift passage’, has been described as a ‘suitable starting point for the navigation of the lower river’. See Harper, R.P., ‘Excavations at Dibsi Faraj, Northern Syria 1972–1974. A preliminary note on the site and its monuments‘, Dumbarton Oaks Papers, 29, 1975, p. 321, n. 4.Google Scholar

⁸⁶ Muraoka, op. cit., 87.

⁸⁷ Wilkinson, J., Egeria's travels to the Holy Land (Jerusalem, 1981), 113.Google Scholar She only provides scant, general details in her description of Hierapolis as a city of ‘great plenty, rich and very beautifulș. Nor does she supply any further information about the crossing point on the Euphrates which, from the distance quoted, could be Qal'ah Najm. Presumably, Christian traditions associated with both Hierapolis and the hinterland had not evolved, for we can expect that they would have been recorded by Egeria.

⁸⁸ Goossens, op. cit., 175, notes that the patron saints of Hierapolis were Peter and Paul. By the sixth century A.D. the city boasted a church dedicated to St. Peter and also the reputed tomb of St. Matthew. However the greatest fame of Hierapolis derived from St. Simeon Stylites whose feats attracted thousands of pilgrims.

⁸⁹ Kawerau, P., Die jakobitische Kirche im Zeitalter des syrischen Renaissance (Berlin, 1960), 43,Google Scholar notes the components and the distinguishing features of monastic lay-out. Several aspects of the Roman station at Djebel Khaled would support its adaptation to a monastery, not least its location on precipitous cliffs. The walls would also have provided protection and the water cisterns in the acropolis area would have supplied the community's needs. The burial chambers could have served as cells for the monks and also provided hospice facilities as well as being converted into a church.

⁹⁰ Clarke, op. cit., 80. The central role of a chapel in a monastic complex is highlighted by Sachau, E., ‘Edessenische Inschriften’, Zeitschrift der Deutschen Morgenldndischen Gesellschaft, xxxvi, 1882, 151.Google Scholar In his discussion of the ruins of Der Ja'kiib, near Edessa, he says, ‘wenn das Gebäude ein Kloster gewesen ware, so müsste man eine Capelle nachweisen können’.

⁹¹ Canaan, T., Mohammedan saints and sanctuaries in Palestine (Facs. ed. Jerusalem, ),Google Scholar 17, specifies that in large shrines which were visited by many pilgrims,‘a servant lives all the year around in such a sanctuary to guard it’.

⁹² Littmann, op. cit. (1904), p. 39.Google Scholar

⁹³ Pognon, op. cit., Inscriptions 62, 75, 92, 94 and 96 which record restoration works ranging from an entire church (Inscriptions 62,96), to an altar (Inscription 75), a door (Inscription 92) and a tower (Inscription 94). Littmann, op. cit., Inscription 62 dating from the fifth to sixth centuries A.D., which occurred on the lintel of the church at Shekh Sleman in northern Syria offers a variant by recording: ‘I, the deacon Euthalios, wrought this. May everyone that reads (this) pray for me.’ Presumably some restoration works did take place, although they are not specified. Whilst the executor identifies himself, the process which he adopts differs from Pognon, op. cit., Inscriptions 62, 75, 94 and 96 where the names are narrated in the third person after the works are detailed. As such, the inscription from Shekh Sleman appears to be a combination of the opening and closing formulae that occur in Inscription II, perhaps representing the genre in a compressed form.

⁹⁴ Muraoka, op. cit., 86.

⁹⁵ See Gollancz, op. cit., where numerous amulets are headed by a summary of the evils with which one is to contend.

⁹⁶ Ref. n. 47.

⁹⁷ Budge, E. A. W., The Chronography of Gregory Abu'l Faraj (Oxford, 1932) 435.Google Scholar

⁹⁸ Pognon, op. cit., Inscription 74 relates the charming tale of the instant conversion of the pillaging Tatars (Mongols) who came upon the bones of a saint upon their arrival at the monastery and were pacified. Doubtless, the writer of Inscription I may have hoped for the same spirit to prevail, should such a situation arise.

⁹⁹ See Budge, op. cit., 420, which narrates that an old man and his son, in the face of the Tatars’(Mongols’) arrival at Margā, ‘did not wish to flee to the monastery with the rest of the people of that village, but he climbed the mountain on the other side, and he went and hid in a strong and rocky place on the banks of the Euphrates.’ This must have been a common reaction.