THE EARLIEST EXAMPLE OF CHRISTIAN HYMNODY Fron Patristic writings ample evidence can be gathered about the important part which hymn-singing held in Early Christianity, Until recently, however, Early Christian hymnography was known only from documents transmitting the text but not the music, The discovery and publication of a Christian hymn in Greek with musical notation was, therefore, bound to change the whole aspect of studies concerned with the history of Early Christian music. This happened, as is well known, in 1922 when, under No. 1786 of the fifteenth volume of the Oxyrhynchus Papyri, A. S, Hunt edited a fragment of a hymn, dating from the late third century, together with a transcript of the music by H. Stuart Jones. For the first time it became possible to realize what kind of music Greek-speaking Christians in Egypt sang in praise of the Lord. ‘The importance of the discovery clearly appears from Hunt's concise and masterly commentary on the hymn (pp. 21-3 and 25), though he rightly refrained from drawing far-reaching conclusions from the strip of papyrus, containing only the last lines of a hymn, four out of five of which are disfigured by large initial lacunae. Such restraint was not observed by all the scholars who commented upon the hymn.? All the questions, however, raised in connexion with the discovery of a document of such outstanding significance could not be solved at once. We soon find three scholars at work, adding new data to the commentary of Hunt, and making some valuable suggestions: Th. Reinach, R. Wagner, and H. Abert. Yet their attempts are not satisfying as far as the restitution of the melodic line is concerned. In his endeavour to present the reader with a coherent piece of Greek music, Reinach sup- presses in several places the lacunae both in the words and in the music, Stuart Jones and Abert do not go as far as Reinach, but they, too, reduce in some places the spacious lacunae in the music to short rests, thus giving an incorrect picture of the original musical structure of the hymn. Wagner's transcript is correct; but his rhythmical interpretation is prejudiced by metrical theories imposed on the music. In this article I shall give a survey of the textual readings of the music suggested by these scholars. I shall then try to show that the very important question of the musical rhythm has been treated as if it were determined entirely by the metre of the text, whereas it can be seen from the papyrus that the singing of the hymn was regulated by signs additional to the musical notation proper. A facsimile of the original strip is shown in Plate I of the O.P., enabling us to compare the musical signs on the papyrus with their rendering into modern notation by Stuart Jones, Reinach, Wagner, and Abert. Hunt’s transcription of the five lines runs as follows. I add to the exact rendering of his text only the figures introduced by R. Wagner in the plate accompanying his study on the hymn, in order to facilitate references to the musical signs: ¥ Oxyrhynchus Papyri, Part xv (0.P.), p. 22. 2 CE. W. Vetter’s exaggerated conclusions in Pauly’s Real-Encycl. d. cl. Alt., ed. W. Kroll 41935), ¢- 874. 2Th. Reinach, ‘Un ancétre de Ja Musique d’Eglise’, Reoue Musicale, July 1922, p. 24, and La Musique grecque, 1926, R. Wagner, ‘Der Oxyrhynchos-Notenpapyrus’, Philologus, Ixxix (NF. xxiii), 1924, pp. 201-215 the transcription is on a separate table. H. Abert, ‘Ein nevent- deckter frithchristlicher Hymnus mit antiken Musiknoten’, Zeitschrift fiir Musikwissenschaft, iv, 1921-2, p. 527, and ‘Das alteste Denkmal der christlichen Kirchenmusik’, Die Antike, i, 1926, pp. 282-0. A reprint of this article in Gesammelte Schriften und Vortrdge von Hermann Abert, herausgegeben von Friedrich Blume, 1929, contains a revised version of the transcrip- tion on p. 89. ‘The example given in Th. Gérold’s Les Péres de VEglise et la Musique, 1931, p. 45, represents an inexact rendering of Th. Reinach’s version of the hymn.THE EARLIEST EXAMPLE OF CHRISTIAN HYMNODY 35 (we ‘Bi letters a covey mata te Peov Aoyypoe i234 67 @ Gioun is 4 (Qt ae letters To boeepa Fk TF luw7o oe yata pd” i 7 it ug zo ai (x) a3 Bee Sr ige cat aeren Gaergoea Aine i 2345 678910 uname Fisend fo Ehap 8% so Totapey godiay Tasas vps (3) tejtov]-Aee-LoT gl ye aap wR eR vovytay quay F284 56 7 egieun nae Tey 2m 2 177 RR ore DIA OF eh F (4 tmerega xonor Xv ayoy mrevpe Tao Syvapes Bon aS by AE ages warIea7 ae ay te an oo FB F USE HE Frgn Foy aS ETGavovITaY anny apy Keates axLv0E 204667 Cgoran as 6 y wg wal NONE eH ywH ED = eee ee GOR woo$ io §F Bt LF ope Osa) sree Silay wyya terray qatar Suqy = 33 an ae 86 of oF & way36 E. J. WELLESZ Before analysing the musical signs I give the transcript of Stuart Jones, who, to judge from a remark of Hunt (0.P., p. 23), seems to have taken an active part in solving some of the problems of the musical notation: Jv tay 4 eo we to pd? Jo-teo ee Gan sett fy) g ti tt 2 ~€6- O5-pa NMarlé- ow... no-ta- por po-O-wy Ta- ~ ~ oar Du- viv toy P- poy ma-té- ea yvi- (2) ™~ oy xe-y- ov -. myep- WO, 0 = Va Les, -m-w- yooy- toy 4 ~ PY a py Kedtos 3 oO Yos ++. Sa- tH- ee Md- Yep ~~ Ft O22 mdy- tar d-ya-Oay 2 — pyY &— HY. The melody has a compass of eight tones, designated by eight letters: Rdaofile. These letters, standing for musical notes, are to be found in the Diatonic Hypolydian key of Alypius,' the mode being the Hypophrygian.2 Thus we get according to F. Bellermann’s interpretation’ the following row of tones: Reso be he ———— == Apart from the signs, indicating musical notes, five additional signs are used by the scribe of the papyrus, regulating the rhythm and the execution of the melody. These signs had already been discussed by Reinach, Wagner, and Abert when com- 1 C£.C. Jan, Musici ScriptoresGraeci (Teubner), Berliner Notenpapyrus’, Philologus, xxvii (N.F. P. 370 xxxi), 1922, pp. 256-310, and H. Abert, ‘Der neue 2 0.P.,p. 22. The same notation occurs in the griechische Papyrus mit Musiknoten’, Archiv first piece of the rather earlier papyrus pub- f. Musikwiss. i, 1918, pp. 313-28. lished by W. Schubart in Sitsungsber. preuss. 3 F, Bellermann, Die Tonleitern u. Musik- Ahad., 1938, pp. 763-8. Cf. R. Wagner, ‘Der _noten d. Griechen, 1847, plate 1.THE EARLIEST EXAMPLE OF CHRISTIAN HYMNODY 37 menting on the Berlin papyrus, where the same signs are also to be found, These signs are: (2) —A horizontal stroke above one, two, or three notes, lengthening their dura- tion. Similar signs are found in Western Plain Chant MSS. and are used in modern editions of Gregorian melodies, in order to mark the lengthening of a single note or a group of notes. (2) v This sign, the Ayphen, binds two or three notes together; it corresponds to the slur, or Zigato sign, in our modern stave notation. (3) 9 The leimma is the sign for the xpévos xevds to which Aristides Quintilianus refers.! It stands for a rest, which can be lengthened by a horizontal stroke. (4) : The significance of the colon is not quite clear, and has been widely discussed. It is obviously a sign of division, relating, however, only to the music, not to the metre of the text. Hunt's rendering of the colon in 0.P. is not always correct. The sign is never set in the middle of the space between words or syllables, but always at the beginning of a word or a syllable. Reinach’s suggestion, therefore, that the colon marked a prolongation of the note pre- ceding it is untenable: the sign indicated a short interruption of the melodic flow. I entirely agree with O. Schroeder’s suggestion that the colon marks an indication for the singer to take breath. In modern notation it can best be rendered by ’, which we use for the same purpose. (5) * The dot is placed either above a note, or, if the note is lengthened, above the horizontal stroke, referring to one or two notes. It indicates arsis. This explanation of the sign is given already by Hunt, who bases his view on a passage from Anon, Bellermann (3. 85) which has been widely discussed by ‘Wagner in his study of the Berlin papyrus* and by other scholars. In our transcription we shall render the dot by a small vertical stroke, placed above the notes or rests to which it belongs, this sign being used in modern Plain Chant editions to designate the ictus. On comparing Hunt's transcript of the Hymn to the Holy Trinity with the facsimile on Plate I, I found that he had omitted a few signs of which only traces can be seen. Most of these divergences had already been noted by R. Wagner, who in some cases suggested two different readings. I, therefore, shall give a list of the signs Hunt has omitted, adding a W in brackets to those which I restore to the text in accordance with Wagner. The figures refer to those placed above the letters in my transcript. Second line x. Hyphen omitted, linking together the missing note with £: traces of a hori- zontal stroke clearly visible. Following Hunt’s suggestion, Reinach supplies in the lacuna in the text [od zdv Se/Aay o]8 ray 7 4-5. Hyphen missing (W). 6-7. Hyphen missing (W). 11-12. Dot above «. 20. Dot above t, not a stroke (W). ar. Trace of a dot above $ (W), 22, Traces of a colon in front of the missing note. Third line 3- The colon consists of a dot and a small horizontal stroke. 7-8. Remains of a { are visible between 7 («) and (8) ({). Wagner has seen them too, * Arist. Quint., ed. A. Jahn, p. 27. 6. ® Berl. Phil. Woch. xlii, 1922, p. 323. 3 See Philol. Ixxvii (N.F. xxxi), pp. 297 ff.38 E. J. WELLESZ but seems to have hesitated to accept a group of two notes on the short syllable fo. 8. Above £ is a stroke, not a dot. 18-19. The dot is set clearly above the second note of the group. ax. The colon consists of a dot and a stroke. Fourth line 5-6. The papyrus shows no traces of a horizontal stroke, which Wagner adds in brackets. 7. The colon consists of a small stroke and a dot. 12, Hunt reads £ Wagner suggests either £ or ¢. Reinach and Abert follow Hunt. From the musical point of view both readings are equally possible and good. A close examination of the facsimile makes it rather improbable that the letter was a ¢: I therefore read é. ax. Hunt writes: ‘The note £ above a of Suvdyets is very uncertain.’ Wagner suggests 0, a solution more satisfactory from the musical point of view. The round hole in the papyrus exactly takes up the space of an o. 28, 0 above the last syllable of émpuvovvraw is a misprint, corrected by Hunt himself (W); the letter is an «, as Wagner first suggested. 31-3. The horizontal stroke covers all three notes of the group. 34-3. Hyphen missing. 38-9. Hunt assumes that after each of the two notes a sign is missing. However, no traces of such signs are left in the papyrus, and from the musical point of view there is no need to interpolate any notes. 40-1. Horizontal stroke and hyphen omitted. Fifth line 1-2. After each o we see traces of missing letters. The vertical strokes are longer than necessary for a single letter; each obviously extended over a group of two signs. 3. The colon is omitted. 4-5-6. Hyphen links together the second and third note of the group, not all three. 7. The stroke above ¢ is omitted. 8. A dot should be added. 18-19. The stroke should cover both letters. 26-7-8. As in the former group (4-6) the hyphen links together only the second and third notes of the group. The stroke, however, covers the whole group; the dot is above £. 29-30-1. The stroke covers all the three notes (W). 33-4. Dot and hyphen are omitted. The stroke must be extended over both signs. 35-6. The dot is omitted. Wagner adds a hyphen. The stroke must again be extended over both notes. From this list we can see that a considerable number of corrections have to be made in Hunt’s text, most of them concerning the rhythm of the music, in order to give an exact rendering of it into modern notation. As the greater part of these additions were made by Wagner, we should expect to find them introduced in his transcription of the hymn. Glancing over his transcription and those by Stuart Jones, Reinach, and Abert we are surprised to find that none of these scholars has taken the trouble to introduce all the rhythmical signs of the text and to render them by corresponding signs of our modern notation. Taking account, for the moment, only of the rhythmical signs in Hunt's text, we find that Stuart Jones, Wagner, and Abert have put in only theTHE EARLIEST EXAMPLE OF CHRISTIAN HYMNODY 39 hyphen, rendering it by a slur, and the rest sign ; they have ignored the dot, the colon, and the vertical stroke, Reinach tried to render the colon asa lengthening of the note preceding the sign; but in doing so was forced by his own preconceptions to shorten the rhythmical value of the subsequent note. Unfortunately he neglected the hyphens. In order to demonstrate his method I shall give a short example of Reinach’s transcription from his article in the Reoue Musicale, followed by renderings of the same phrase by Wagner and Abert: To-ta- Gr a go-8- a To-ta- wav Pe-h- oY oe ADD To-ta- May fO-B- OY TA- Tu = =e z mo-ta-paY Go-M- ay Td~ Car Reinach’s version makes a very difficult reading from the rhythmical point of view, and the difficulties are increased by the complete lack of phrasing signs (hyphens). The rhythmical interpretation of the colon is too artificial to be accepted as a satisfactory solution. Wagner's version, on the other hand, is correct, but gives no impression of the flow of the melody. The same is true of Abert’s first version, which probably served Wagner as a model.* The introduction of bars into composi- tions of an age which had no notion of them is a misconception, aggravated, parti- cularly in Wagner’s version, by the fact that the time-unit is indicated by a crotchet instead of a quaver; which makes the melody sound like a nineteenth-century hymn- tune, In his frst transcription of the hymn Abert, too, has introduced bars, but he took the crotchet as time-unit. In his second version, however, he no longer breaks up the flow of the melody with bar-lines, which unconsciously force the singer to treat each group of notes separated by these strokes as being of equal duration. I cannot enter into the theory of this question more fully. I only want to say that this pre- * Reinach himself was not certain that his ‘suggestion would solve the question and worded it rather cautiously: ‘La seule explication que je puisse concevoir de ce signe mystérieux, c'est @y voir une tenue, cest--dire Panalogue du point de niveau... . Je ne donne cet essai d’interprétation que sous toute réserve’. Revue Mus., p. 21. 2 Cf, footnote 2 in R. Wagner's article in Philol.y p. 201. Abert, setting the hymn in the key of Gi minor, and Wagner, in AD minor (its equivalent), are following H. Riemann’s deduc- tions in his Handbuch d. Musikgeschichte, i (1904), p- 198, showing that the Dorian mode in A minor was the Greek fundamental scale, contrary to Bellermann’s hypothesis that it was the Hypolydian. From the theoretical point of view Abert and Wagner were right in setting the hymn in the key of G# minor or Ab minor. In practice, however, there is no need to write the melody in a key which even in modern music is rarely used. In his second version Abert came back to the simpler way of transcribing used by Stuart Jones.40 E, J. WELLESZ conceived metrical conception has hitherto prevented all scholars who have dealt with the hymn from taking account of the rhythmical signs which the text in the papyrus contains. According to Hunt the poem is written in anapaests, a metre which seems to have been a favourite one with Early Christian hymnologists in Egypt. Hunt speaks of ‘anapaestic dimeters, either acatalectic, catalectic, or brachycatalectic’, handled in a free manner, asa short syllable is allowed to replace a long at the end of a colon, and the first syllable of dyrfv is lengthened metri gratia.? A similar view is expressed by Th. Reinach. He draws attention, however, to metrical irregularities in the third phrase of the fragment (third line, 16, to fourth line, 15) Sprotvren 8 judav rarépa xuidv xdyiov med pa. which show in his opinion that the poet makes use of a classical metre which he can no longer master? R. Wagner explains the metre differently. He does not see irregularities in it, but assumes a change from anapaests to catalectic dactylic heptameters.3 Wagner's metrical interpretation is based on his view of the coincidence of the short syllables with musical signs above which a dot is placed, indicating ictus. There are, however, some exceptions to be noticed, and Wagner is forced to add dots in brackets. Wagner's theory was rejected by K. Muenscher, who points to the tendency of poets of that period to introduce new metres or to revive old ones. According to Muenscher the third phrase is a hypercatalectic anapaestic trimeter. (Muenscher counts the rest as a hypercatalectic syllable.) He therefore considers the metre of the fragment as a whole as a mixture of acatalectic dimeters and hypercatalectic trimeters.* From this survey the following conclusions can be drawn. All scholars dealing with the metre of the hymn agree that it is built on an anapaestic system but that some of the verses are handled in a free way ; even the introduction of dactyls has to be accepted in order to get a proper metrical scheme. After considering these various interpretations we may think of another solution of the difficulties with which we are presented in trying to read the hymn as if it were an example of Greek classical poetry. Looking for analogies in other Christian poetry of the same period it seems to me that Reinach, Wagner, Muenscher, and Abert were biased by comparing the fragment with hymns of Clement of Alexandria or Synesius. ‘These hymns, written in an archaizing, strictly maintained anapaestic metre, aimed at continuing the line of Greek classical poetry. For the fragment we may find more suitable analogies in the group of anonymous hymns from the Anthologia graeca carminum Christianorum.s Indeed, the following lines from the evening hymn Aivetre, maiSes: Zoi mpéner alvos, ool mpérer tuvos, ool 83fa mpéret 16 narpi kai 76 vid kal 78 dyiyp mvedpart els robs alvas rév aldésvev. duy, or the sunset hymn és ‘Aapé Sores er Tv apNov Bow dures ais éomepwdv Spvodper rarépa, vidv ral Eyiov nvedjue. Bedy, 1 O.P.,p. 22. 4K. Muenscher, ‘Zum christl. Dreifaltig- 2 Reo. Mus. p. 17. keitshymnus aus Oxyrhynchus’, Philol. lxxx 3 ‘Zu beaeichnen ist sie als heptameter (N.F. xxxiv), 1924, p. 212. dactylicus catalecticus in syllabam.’ Philol., 5 W. Christ and M, Paranikas, Anth, gr. carm. p. 208. Christ. 1871, pp. 38-49.THE EARLIEST EXAMPLE OF CHRISTIAN HYMNODY 4 are more closely related to the third passage from the papyrus than any verse of Clement or Synesius. The lines from the two hymns, both dating from the pre- Constantinian period, are also doxological formulae, taken either word for word or in slightly altered form from hymns of praise of the Early Christian Church. This kind of poetry can be traced back to the days of the Primitive Church, and from there to the hymns and psalms of praise of the Jewish liturgy. All these invocations are composed in an exalted rhythmical prose, characteristic of many hymns of that age. The same can be said of the other passages of the hymn, which cannot either be considered as original. We have to bear in mind that the work of the Christian hymn- writers is difierent from that of a classical poet. All the hymns were sung. The com- position of a new hymn consisted in adorning passages taken from the psalms or ‘songs of praise by the addition of some new passages, or even only a few words, and singing this new text to the melodic phrases of the original chant which was well known to all. The adaptation to the new text was achieved by adding a few notes or interpolating a cadence in conformity with the structure of the chant.2 The metre of the text of the fragment which we are examining, therefore, is not the result of the archaizing tendency of an individual poet but of elevated diction to which the Hellenistic hymn-writer was accustomed from other hymns of the service. His task was either to translate a Jewish or Syrian hymn into Greek, or to write a new hymn on the pattern of an older one. Setting the words to music he tried to write anapaests, the popular metre of the Hellenistic age. But when he came to insert the doxological formula, the wording of which could not be altered in more than slight details since its text, prescribed by the liturgy, was sung to a stereotyped cadence, he had to abandon the anapaestic metre and to introduce rhythmical prose. From these considerations let us turn back to the music, the examination of which shows that its rhythm has no longer that intimate, if not rigid, connexion with the metre of the text which was characteristic of the Greek classical style of composition. Metrically short syllables are often set to lengthened notes; and, in an age when spoken Greek was becoming a stressed language,’ the accentuated syllables are not consistently related to the musical ictus. The notes which have to be lengthened are marked by vertical strokes, and notes which have to be accentuated are distinguished from others by a dot. The rhythmical scheme, adopted by modern scholars in their renderings of Classical Greek music into our stave-notation was to make a note above a long syllable equal to a crotchet, and, correspondingly, a note above a short syllable equal to a quaver. We are not entitled to apply this scheme to the music of the Christian hymn. For in the Christian hymn we have to deal with rhythmical nuances which are too subtle to be expressed by doubling the time-value of the note to be lengthened. We must write all notes as quavers and indicate by an episema (a term known from Gregorian Chant, i.e. a horizontal stroke), those notes which have to be lengthened. ‘Adding the signs omitted in the first edition by Hunt and Stuart Jones, we get the following transcript of the text of the papyrus, and its rendering into modern musical notation: * Cf. A. Baumstark, Liturgie Comparée, 1939, pp. 69 ff., and Th. Gérold, Les Péres de Péglise et la musique, 1931, pp. 19 ff. 2 For a more detailed discussion on this Principle of composition I may refer to my study ‘on ‘Eastern Elements in Western Chant’, Mon. Mus, Bys., Amer. Ser. vol. i (in print). 2 An example of this state of transition is given by C. Wessely in his ‘Les plus anciens monuments du Christianisme’, Patr. orient, iv, Pp. 205 ff., commenting on the metre of the Christian hymn (third or fourth century) from the Amherst Papyri, i, ed. Grenfell and Hunt, pp. 23-8: ‘La construction métrique a pour base deux principes : 'un est celui de 'ancienne poésie grecque, l'emploi alternatif de syllables Tongues et bréves; l'autre est l'accentuation des syllables.’42 E. J. WELLESZ (NC 3 letters a enov Maca t6 Peov Aoyymor ate) Glagl +: 1 a2 ur sig 13 tb ep apn 2 SUS (iL 28 letters - jy ty _7- Qornn w Sete a fi io ne @ ke ante fend $t tit ee 2 evs yd- tw pal? B- orga go-€-eps-pa —Aferte ves] No- ta-pay go- O- cor 24 R rE- gar. b-pvow-toy 34 — way S6é 769 wep now 1 6 ee 789 ,aa maou wu rag FH Roe o Ob en] 3 (iv) [nJa- ré-ea yri-dy Y&- y- ov aved-ma.THE EARLIEST EXAMPLE OF CHRISTIAN HYMNODY 43 Tg 20 we az 2B a aS 2827 o> ood 005 oe wt w ” son n@~ Gar Sy-va-pes €- me pw-vooy-tar” &- wHY nds 5 Be & pny Ked- tos, B= vos 1 2 3856 7 fgen 2m 67 ) by 108 RD MF BB yw 3 3s 56 EE op eh ooh Fy oe oF With the exception of the Paean from the Berlin papyrus, dating from the end of the second century A.D. or the beginning of the third, no other piece of Greek music has come down to us which has so rich a flow of melody as that of the hymn in praise of the Holy Trinity. Abert speaks of its genuinely Greek character," and takes the hymn as a proof that Greek pagan influences as well as Jewish characteristics can be traced in Early Christian music.? Wagner, though agreeing in principle with Abert, points to certain -similarities between cadences in Gregorian melodies and those in the Christian hymn ;* but he warns us to regard the hymn as a pastiche, since its personal style (he declared) showed that it was intended for private devotion.+ * Zeitschr. f. Mus. Wiss, 1921-2, p. 529. # Ibid. 528, At the time when these statements were made only a few Byzantine melodies 3 Philol., 1924, pp. 213-14. 4 Ibid, p. 213.4 E, J. WELLESZ had been deciphered, and little was known in general about the style of Early Byzan- tine music. At present, however, we not only possess a fairly complete survey of the rich treasury of Byzantine melodies, but we also know that the shape and character of these melodies had changed little during the period from which manuscripts are preserved, i.e. from the tenth to the fourteenth century.? The kernel of the melodies of the Hirmologium, undoubtedly the oldest layer of the Byzantine corpus of hymns, dates from an even earlier stage of hymnography, namely from the age of the Melodi, Romanus, and the other writers of Kontakia from the early sixth century, who translated the poetical homilies of Syriac hymn-writers into Greek,’ and adapted the melodies of the poems to these new texts. Evidence for the Syro-Palestinian origin of the Hirmi can be gathered from the fact that they show a principle of composition everywhere to be found in the Middle East,* but unknown in Greek music: the con- nexion of certain melodic formulae, linked together by varying short passages in the manner of a recitative. The repetition of some melodic formulae in the hymn was already noticed by Abert and Wagner, but at the date of the publication of their article the significance of the fact could not be recognized. The addition and combination of certain formulae, characteristic of a mode, was not yet discovered as a principle of composition in the music of the Middle East. Together with the new creed and the chants and hymns of Early Christianity this musical style penetrated to the countries of the Mediter- ranean basin. Thus it is not surprising to find similarities between some cadences of the Christian hymn and Gregorian antiphons. But the relation between typical musical phrases in the Hirmi and some of the formulae of the hymn are even more striking. The melodic phrase to the words warépa yvidv ydysov mvedpa (fourth line, 1-16) and its repetition (fifth line, 1x4) are cadences typical of Byzantine music, or, more correctly, combinations of two Byzantine cadences which frequently occur in Hirmi of the fourth Mode, as can be seen from the following examples, taken from Codex Iviron (c. 1130)® (the structural notes of the cadence are marked by asterisks): fol. 60v. SSS Xet-pas Ek-te- ta- cay Aa-ve-nr fa. itty. ’ cee eee == SS Aeyrrtes 0 Twv. Ta-TE-pUY > RAOV * Cf. E, Wellesz, Trésor de Musique Bysan- tine, i, 1934, and Monumenta Musicae Byzantinae, (Pranseripta, i (1936) and ii (1938), ed. C. Hoeg, Hi. J. W. Tillyard, E, Welles). 2°, Wellesz, ‘Ober Rhythmus u. Vortrag d. byz. Melodien’, Bys. Zeitschr. xxxiii (1933), pp. 62-6. ‘Studien z. byz. Musik’, Zeitschr. f. Mus. Wiss. xvi (1934), pp. 217 £. 2 C, Emereau, Saint Ephrem le Syrien, 1918, Pp. 103 fl. + Cf. my article on Eastern Church Music in Grove’s Didionary of Music, Supplem. Vola, 1940. 5, Wagner bases his deductions in Philol., loc. cit. on A. Gevaert's La Melopée antique, published in 189s. We cannot accept the material used by Gevaert as a reliable source, because the melodies are not based upon the MSS. of the best period, such as those published by the monks of Solesmes. Moreover, the Gregorian melodies underwent many transformations when the Roman rite crystallized, and features characteristic of the pre-Gregorian period were destroyed. © Hirmologium Athoum, ed. C. Héeg; Cod. Monast. Hiber. 4703 Mon. Mus. Bys. Facs., vol. ii, 1938.THE EARLIEST EXAMPLE OF CHRISTIAN HYMNODY 45, To the final cadence of the hymn éyjv, dyjy (fifth line, 26-36) the closing formula of Hymn 1603 (Cod. Ivir. 131 r.) provides a fitting parallel : . wae * a ey Tus ~ we vas wae > te Daeg vpov-Te av~ Toy kaw The cantillation on 4 natural to the words (8v)-rdyes émpuvorr-(rwr) (fourth line 21-6) is a good example of the linking together of two melodic phrases. Its initium is on g (Sv-), in order to link up with the preceding phrase ending on the same note; while its close leaps up to d (-ra»), in order to prepare the beginning of the next phrase (dv) on e. This kind of cantillation is typical of the Early Christian liturgical singing, derived from the singing of the psalms in Jewish liturgy ; it came down to us virtually unchanged, both in the practice of Jews in the Middle East and in the Eastern and Western Churches." From these considerations, based on stylistic analysis, the view can no longer be maintained that the music of the hymn is of genuinely Greek character, nor can we follow Abert? and Wagner? in regarding it as the outcome of a kind of ecstatic im- provisation. In spite of its Greek notation‘ the hymn is an example of the new kind of ecclesi- astical music, modelled on patterns deriving from Oriental sources and used for the Greek text. In transcribing and performing the hymn we are not, therefore, permitted to introduce a rhythmical scheme which may be too rigid even for melodies deriving from the classical period of Greek music. We have to take into account and to reproduce the rhythmical signs from the papyrus, which would not have been used in such abundance if the metre of the text had been thought sufficient to regulate exactly the execution of the music. In interpreting the hymn according to the principles of musical palaeography we arrive at a melody the structure and expression of which already show the features characteristic of the treasury of Byzantine ecclesiastical music. E. J. Wettesz, ¥ A. Z. Idelsohn, ‘Parallelen zwischen gregor. u, hebr.-orient, Gesangsweisen’, Zeitschr. f. Mus. Wiss. iv, 1921-2, pp. 514-24. 2 Zeitschr. f. Mus. Wiss. iv, p. 528. 3 Philol., loc. cit. p. att. * Like all other Oriental music, Early Chris- tian music both Eastern and Western was transmitted orally. The signs of Early Byzan- tine notation originally regulated the rhythm and execution. Interval signs were introduced only by degrees, when it became increasingly difficult to store all the melodies in the memory. ‘We may, however, well imagine that a Christian in Egypt, trained in music, may have used the system of notation handed down by Alypius, in order to remind him of the hymn.