1 I am grateful to the National Endowment for the Humanities and to the American Academy in Rome for their support of the research reported hereGoogle Scholar

2 Chronica monasteni casinensis, ed Hartmut Hoffmann, Monumenta germaniae historica, Scriptores, 34 (Hanover, 1980), II, 94Google Scholar

3 On Frederick’s career as related to Montecassino, see Herbert Bloch, ‘Monte Cassino, Byzantium, and the West in the Earlier Middle Ages’, Dumbarton Oaks Papers, 3 (1946), 189-93, and Jules Gay, L’Italie méridionale et l’empire byzantin, Bibliothèque des Écoles Françaises d’Athènes et de Rome, 90 (Paris, 1904), 509-61 Frederick had opportunities to hear the ‘Ambrosian’ chant not only at Montecassino, but also at Benevento and Tremiti See my ‘Montecassino and the Old Beneventan Chant’, Early Music History, 5 (1985), 53-83 (pp 79-83)Google Scholar

4 Henry Marriott Bannister, ‘Ordine “Ambrosiano” per la settimana santa’, Miscellanea ceriani (Milan, 1920), 127-41, Myrtilla Avery, ‘The Beneventan Lections for the Vigil of Easter and the Ambrosian Chant Banned by Pope Stephen IX at Montecassino’, Studi gregoriani, 1 (1947), 433-56Google Scholar

5 The recent article by Terence Bailey, ‘Ambrosian Chant in Southern Italy’, Journal of the Plainsong and Mediaeval Music Society, 6 (1983), 1-7, is the first serious study of musical relations in the two liturgies. See also Ambros Odermatt, Ein Rituale in beneventanischer Schrift (Freiburg/ Schweiz, 1980), 66-9, Bonifacio Baroffio, ‘Le origini del canto liturgico della chiesa latina e la formazione dei repertori italici’, Renovatio, 13 (1978), 44-6, [René-Jean Hesbert et al], Paléographie musicale, xiv (Tournai, 1931, repr Berne, 1971), 453-6, John Boe, ‘A New Source for Old Beneventan Chant The Santa Sophia Maundy, in MS Ottoboni Dat 145’, Acta musicologica, 52(1980), 122-33Google Scholar

6 Elias Avery Loew, The Beneventan Script, 2nd edn, rev Virginia Brown, 2 vols, Sussidi eruditi, 33-4 (Rome, 1980), 1, 75, indicates that the manuscript bears an ex libris from the Cassinese dependency of S Maria de AlbanetaGoogle Scholar

7 The text was published, but imperfectly, in Ambrogio Amelli, Paolo diacono, Carlo Magno, e Paolino d’Aquileia in un epigramma inedito intorno al canto gregoriano e ambrosiano (Montecassino, 1899)Google Scholar

8 Facsimile in Paléographie musicale, xiv, Plates XXXII, XXXIII, and in Henry Marriott Bannister, Monumenti vaticani di paleografia musicale latina, 2 vols (Leipzig, 1913), ι, Plate 73. These two pages contain music for Tuesday and Wednesday of the second week of Lent in the Ambrosian rite All the pieces on this leaf are to be found, in the same order, in the Ambrosian gradual-antiphoner London, British Library, Add MS 34209, pp 174—5 (facsimile m Paléographie musicale, v (Solesmes, 1896), transcription in Paléographie musicale, vi (Solesmes, 1900), 204—5)Google Scholar

9 Evidently his model was not in Beneventan notation, the scribe used the ‘Ambrosian’ climacus not found in Beneventan notation See Michel Huglo et al, Fonti e paleografia del canto ambrosiano, Archivio ambrosiano, 7 (Milan, 1956), no 40, pp 21-3Google Scholar

10 See Kelly, ‘Montecassino and the Old Beneventan Chant’Google Scholar

11 The mass begins on f 102^v Int Clamaverunt, Gr Clamaverunt, All Sancti mei, Off Anima nostra, Comm Justorum animae For the Italian Alleluia Sancti mei, see Karl-Heinz Schlager, Thematischer Katalog der altesten Alleluia-Melodien (Munich, 1965), no 273, p 196 The same mass is found in Benevento 38, f 118^v, but with two other Alleluias. Sancte Nazari vir dei (Schlager no 262 — a Beneventan product) and Te martyrum (cued only - Schlager no 397)Google Scholar

12 Antiphonale missarum juxta ritum sanetae ecclesiae mediolanensis (Rome, 1935), p 503, Oxford, Bodleian Library, MS lat. liturg a 4, f 171.Google Scholar

13 The melody is that printed in the Antiphonale missarum, pp 607-9, as ‘Tonus Solemnior’, for Ambrosian manuscript sources of this Gloria see the list in Huglo, Fonti e paleografia, 241Google Scholar

14 Gr Anima nostra (cued), All Sancte Nazari, All. Sancti mei, Seq. Laetetur celum humus, Sanctus with Osanna plasmatum populum, Off Anima nostra (cued). Comm Justorum (cued)Google Scholar

15 These are studied in detail in Thomas Forrest Kelly, ‘Non-Gregorian Music in an Antiphoner of Benevento’, The Journal of Musicology, 1987 (forthcoming)Google Scholar

16 The correspondence may of course be accidental Naturally when a book is overlarge it is convenient to divide it, as is often done with lectionaries and breviaries But this is not the only Beneventan book arranged in this way Vatican lat 10645, a collection of fragments taken from bindings, includes as f 63 a single leaf from a combined missal-breviary with musical notation of the twelfth century With a single leaf we cannot say whether the book was divided (though it would probably have needed to be), nor whether the dividing point was at Easter The scribe of Benevento 38 seems to have used a Beneventan ‘totum’ combining mass and office as his source for Beneventan melodies in the old Beneventan masses many pieces with texts specific to their feast are cued, doubtless matching the scribe’s exemplar, which must have abbreviated the pieces in question because they also occurred in the office portion of the same manuscriptGoogle Scholar

17 Arguing against this, however, is the fact that the flyleaf (f 202) of Benevento 35, from a Beneventan gradual, uses only one Communion for Christmas The case would be stronger if, in cases where there are two Communions, the text of the second were similar to the texts of many Ambrosian transitoria, centering on access to the altar for Communion (Accedite ad altare Dei, Venite, convertimini ad me, Accedite et edite, etc)Google Scholar

18 See Terence Bailey, The Ambrosian Alleluias (Egham, Surrey, 1983), esp pp 46-52, 88-91 Some Ambrosian melodies are elaborations of other melodies, thus reducing the original number still further Bailey believes that the frequently used Beneventan melody is related to an Ambrosian melody, see his Example 7(a), p 53Google Scholar

19 For the Milanese Gospels, see Angelo Paredi, Sacramentarium bergomense, Monumenta bergomensia, vi (Bergamo, 1962), 100, 109-10, 119-20, 129-30, 137-9, Benevento 33 has these same Gospel pericopes (see the facsimile of Benevento 33 in Paléographie musicale, xx (Berne and Frankfurt, 1983), and Sieghild Rehle, ‘Missale Beneventanum (Codex VI 33 des Erzbischoflichen Archivs von Benevent', Sacris erudiri, 21 (1972-73), pp 343 (no 44), 344 (no 51), 345 (no. 58), 346 (no 65), 348 (no 72) The same lections are to be found in a thirteenth-century missal of Ragusa, Oxford, Bodleian Library, MS Can liturg 343, see René-Jean Hesbert, ‘Les dimanches de carême dans les manuscrits romano-beneventains’, Ephemerides liturgicae, 48 (1934), 198-222Google Scholar

20 Studied in Alban Dold, ‘Fragmente eines um die Jahrtausendwende in beneventanischer Schrift geschriebenen Vollmissales aus Cod Vat lat 10645’, Jahrbuch für Liturgiewissenschaft, 10 (1930), 40-55Google Scholar

21 Facsimile in Paléographie musicale, xiv, Plate VIIIGoogle Scholar

22 Among the Gregorian masses there are two Communions for Lent II Redimet (common to Beneventan manuscripts), and Qui biberit, referring to the Gospel of the Samaritan woman The same two Communions are cued for Friday of the second week of Lent, evidently there may be a choice of pericopes, the coming together of separate traditionsGoogle Scholar

23 See Edmond (Eugène) Moeller, Corpus praefationum, 5 vols, Corpus christianorum, series latina, vols CLXI, CLXIA-D, (Turnhout, 1980-1), vol CLXI, pp LXXXI-LXXXIII, the shared Prefaces for the Sundays in Lent are nos 502, 627, 799, 920 and 1594 in Moeller’s catalogue Additional Prefaces from Benevento 33 used also at Milan include Moeller’s nos 616bis, 637 and 1051Google Scholar

24 Klaus Gamber, ‘Die kampanische Lektionsordnung’, Sacris erudiri, 13 (1962), 344—5, and ‘Das kampanische Messbuch als Vorläufer des Gelasianum Ist der hl. Paulinus von Nola der Verfasser?’, Ephemerides liturgicae, 12 (1961), 31-42Google Scholar

25 Klaus Gamber, ‘Die Sonntagsmessen nach Pfingsten im Cod VI 33 von Benevent’, Ephemerides liturgicae, 74 (1960), 428-31Google Scholar

26 Jacques-Paul Migne, Patrologia latina, xvi (Paris, 1880), col 1044Google Scholar

27 See Paléographie musicale, xiv, 271—4Google Scholar

28 Vatican lat 10673, Benevento 33, 40, 39, see Paléographie musicale, xiv, 318-21; René-Jean Hesbert, ‘L’“Antiphonale missarum” de l’ancien rit bénéventain’, Ephemerides liturgicae, 60 (1946), 103-41 (pp 120-6), Paléographie musicale, v, 248-50Google Scholar

29 René-Jean Hesbert, ‘Le Répons “Tenebrae” dans les liturgies Romaine, Milanaise et Bénéventaine contribution à l’histoire d’une interpolation evangélique’, Revue grégorienne, 19 (1934), 4-24, 57-65, 84-9, and ‘L’“Antiphonale missarum’”, Ephemerides liturgicae, 60 (1946), 127-133, also Paléographie musicale, xiv, 326-31Google Scholar

30 Marcus Magistretti, Beroldus, sive ecclesiae ambrosianae mediolanensis kalendarium et ordines saec XII (Milan, 1894), 106-7 This double ceremony is attested in most, but not all, Ambrosian books The second Adoration is not mentioned in Milan, Biblioteca ambrosiana, MS T 103 sup, a manual of the eleventh century, I 27 sup (twelfth century) or Y 18 sup (fourteenth-fifteenth century) See Enrico Cattaneo, ‘L’adorazione della croce nell’ antico rito ambrosiano’, Ambrosius, 9 (1933), 181-5Google Scholar

31 Further on these antiphons in the Beneventan liturgy, see below, pp 191-4, Paléographie musicale, xiv, 308-13, and Hesbert, ‘L’“Antiphonale missarum’”, Ephemerides liturgicae, 60, 104—16Google Scholar

32 Cum accepisset, which is used in the Gregorian liturgy also for Good Friday and whose Ambrosian and Gregorian melodies are closely related, had a substantially different melody at Benevento See Paléographie musicale, xiv, 335-7, and Hesbert, ‘L’“Antiphonale missarum’”, 136-41Google Scholar

33 Hesbert, ‘L’“Antiphonale missarum’“, Ephemerides liturgicae, 61 (1947), 153-210 (pp 197-201), Paléographie musicale, xiv, 429-32, and v, 219, Pietro Borella, ‘La “Missa” o “Dimissio catechumenorum” nelle liturgie occidentali’, Ephemerides liturgicae, 53 (1939), 60-110 An additional source of this dismissal, not noticed hitherto, is in Vatican Barb lat 631, a late eleventh-century pontifical of Montecassino On f 75, after the pre-baptismal rite and before the propers of Holy Saturday, is found the single dismissal ‘Si quis catecuminus est recedať, without musical notationGoogle Scholar

34 Oxford, Bodleian Library, MS lat liturg a 4, f 4 ‘Ad cornu altaris dicat presbiter tribus vicibus Xristus dominus resurrexit Deo gratias ‘ Benevento 38, f 53 ‘Deinde dicat sacerdos III vicibus Iam xpistus dominus resurrexit. cleri simul cum populo Deo gratias' See Michel Huglo, ‘L’ annuncio pasquale della liturgia ambrosiana’, Ambrosius, 33 (1957), 88-91Google Scholar

35 Details of usage, sources and textual readings are omitted from Table 2, our purpose here is simply to lay out the extent of the material shared by the two liturgies Biblical texts which vary among the rites have been omittedGoogle Scholar

36 The official ordinal produced by Beroldus early in the twelfth century (ed Magistretti, Beroldus) is probably part of the same effort at codification that produced the earliest of the surviving complete chant books, also from the twelfth century. The principal early Ambrosian manuscripts are listed in Bailey, The Ambrosian Alleluias, 102, for a more complete catalogue, see Huglo, Fonti e paleografia, and the additions noted in Bonifacio Baroffio, ‘Le origini del canto liturgico nella chiesa latina e la formazione dei repertori italici’, Renovatio, 13 (1978), pp 47-8, n 85Google Scholar

37 One such comparison can be made by the reader who has access to vols v and xiv of Paléographie musicale, there he may compare the Canticle of the Three Children (Benedictus es, domine deus patrum nostrorum), with its preface Tunc hi tres and its concluding verses Though the elaborate melodies are not the same, the structures of the two canticles are in many ways closely parallel The introduction and the final verses are alike in each version, though each rite uses its own recitation tone And the melodic structures match in the canticle The opening words ‘Benedictus es [1] domine [2] deus [3]’ have the same melody three times (melody A), though the music of Benevento is not that of Milan And in each rite another melody (B) appears on ‘laudabilis’ and on ‘saecula Amen’ These two melodies function throughout the canticle at parallel places. ‘Ymnum dicite’ in each matches that canticle’s melody B, in two other places melody A is used three times in each ‘Benedicite [1] omnia [2] opera [3]’, and ‘Benedicimus [1] patrem [2] et filium [3]’ And other matching texts have matching music, as for Domini dominum (at Milan, Domini domino) The structure is the same, but the melodies are differentGoogle Scholar

38 Related Beneventan pieces are the Offertory Ascendit deus (Benevento 38, f 93^v, Benevento 40, f 7^v), the Communion Quid ad nos (Benevento 38, f 140^v, Benevento 40, f 143) and the Communion Gaudens in celis (Benevento 40, f 133^v)Google Scholar

39 For a more detailed study of these texts, see Hesbert, ‘L’“Antiphonale missarum’”, Ephemerides liturgicae, 59 (1945), 69-95 (pp 70-3)Google Scholar

40 The Gregorian melodies can conveniently be seen in modern publications such as the Graduale sacrosanctae romanae ecclesiae (Tournai, 1948), 200 and 202Google Scholar

41 Benevento 35, 39, 40, Lucca 606, the first of them only is in Benevento 30 and 33Google Scholar

42 The musical parallel is noted also in Huglo, Fonti e paleografia, 122, and in Kenneth Levy, ‘The Italian Neophytes’ Chants’, Journal of the American Musicological Society, 23 (1970), 180-90Google Scholar

43 Milan repeats the Adoration antiphons for the feast of the Exaltation of the Holy Cross (Magistretti, Beroldus, 126); and a Beneventan mass for the Holy Cross (Benevento 40, f 124^v) cues two pieces (Adoramus and Crucem tuam) for Offertory and Communion that may indicate repetition of the Good Friday antiphons (no music is given) This Beneventan mass, though it is found in Benevento 40 at the feast of the Exaltation of the Cross, is the first of two Beneventan masses there, and may actually be intended for the Exaltation, thus repeating antiphons from Good Friday exactly like its sister liturgy in MilanGoogle Scholar

44 That this series of antiphons did not originate at Benevento has been suggested by Johann Drumbl The central antiphon at Benevento, Crucem tuam adoramus, has a Greek original that is widely used in translation in the Latin liturgies, different versions resulting from the adoption of different portions and versions of the original Greek One Greek text is found in the (Rome, 1887), 227-8 and in the (Rome, 1876), 227-8, a shorter version, now used for Sundays of the Cross in Lent and for the feast of the Exaltation is found in the (Rome, 1879), 362 and elswhere The various Latin versions all begin with the same text ‘Crucem tuam adoramus domine et sanctam resurrectionem tuam (Gregorian laudamus et) glorificamus’ There are various continuations: ‘Venue omnes adoremus christi resurrectionem’ (Beneventan), ‘Quia venit salus in universo mundo’ (Old Roman), ‘Ecce enim propter lignum venit gaudium in universo mundo’ (Gregorian) See Johann Drumbl, ‘Zweisprachige Antiphonen zur Kreuzverehrung’, Italia medioevale e umanistica, 19 (1976), 41-55 The arguably awkward Latin of one of the Beneventan antiphons, on the basis of which Hesbert (Paléographie musicale, xiv, 311 ‘une latinité assez contestable’) argued for the antecedence of the Greek version (Hesbert, ‘L’“Antiphonale missarum’”, Ephemerides liturgicae, 60, 1946, 104-10), has been questioned by Drumbl (‘Zweisprachige Antiphonen’, 45-9), who sees the Beneventan version of this text as derivative of an original version composed elsewhere, perhaps at Ravenna, in which the Latin text, influenced by the vernacular, may be the originalGoogle Scholar

45 Compare the later versions from Ravenna and Nonantola in Drumbl, ‘Zweisprachige Antiphonen’, 50Google Scholar

46 The B♭s in the British Library manuscript were added in a not very consistent way by the use of a green line, possibly by a later and not very artistic hand But the B♭ is confirmed by other readingsGoogle Scholar

47 The reader who wishes to pursue matters on his own (and much research remains to be done) already has many further comparisons available in the pars hiemalis in vols xiv and v of Paléographie musicale, and that series proposes to present in the near future both a complete set of facsimiles of the Beneventan chant and the pars estivalis of the Ambrosian liturgyGoogle Scholar