^{page 126 note 1} Gastoué, , Am. Introduction à la Paléographie Musicale Byzantine (Paris, 1907)Google Scholar; Fleischer, O.Neumen-Studien, T. iii. (Berlin, 1904), 66 ff.Google Scholar; Gaisser, U.Les Heirmoi de Pâques (Rome, 1905), 18 ff. and 88 ff.Google Scholar; Riemann, H.Die byzantinische Notenschrift (Leipsic, 1909), 17 ff.Google Scholar

^{page 126 note 2} The quaver is here adopted as the unit of length in accordance with the practice of modern books on plain-song. For the system of Chrysanthus, where minuter subdivision is frequent, the crochet is more convenient.

^{page 126 note 3} See below, p. 133.

^{page 126 note 4} The interval-signs have the same values in the Round and in the Cucuzelian systems; but the use of rhythmical signs is different.

^{page 127 note 1} Quoted by Thibaut, J.Origine byzantine de la Notation neumatique de l'Église latine (Paris, 1907), 58.Google Scholar

^{page 127 note 2} Many signs are supposed to have been derived from certain gestures made with the hand to indicate various progressions of the melody. Cf. Thibaut, ibid. 56.

^{page 127 note 3} For all symbols connected with rhythm, v. Fig. 1.

^{page 127 note 4} This famous treatise is ascribed by some authors (e.g. Gastoué, op. cit. 21) to the elder Cucuzeles, who may have lived about 1300. It seems likely that the work, originally more ancient, has passed through many hands and that its extant forms are comparatively late editions.

^{page 127 note 5} E.g. the manuscripts Vatican, , Barb. gr. 300Google Scholar, and Laconia, , Ἅγιοι Τεσσαράκοντα, 76 do not contain it.Google Scholar

^{page 127 note 6} Published by Paranikas, in Ἑλλ. Φιλ, Εύλλογος, τόμος ΚΑ′, 164Google Scholar; cf. Fleischer, op. cit. c. 3.

^{page 127 note 7} Another MS. Cambridge Univ. Lib. G.g. 1. 2. F. gives τρεῖς ῆμισυ ἀργ and ῆμυισυ ἀργειαν while Riemann (op. cit. 39) seems to have found τρεῖς ῆμισυ μεγὰλαι ἀργὶαι in his copy. It is possible that the writer here meant to say that there were ‘three and a half’ prolongation-marks:—this would be the meaning in late Greek. The total result in either case is the same. But to say that there are ‘three half prolongation-marks,’ and then add ‘But the Tz. has half prolongation’ is nonsense. Hence the need for the emendation already suggested. Ἀργία, not ἄργεια , seems to be the proper form.

^{page 129 note 1} The identity of Klasma and Tzakisma seems certain. They originally had the shape ν and later υ. The form given by Riemann, op. cit. is unusual. He makes the mistake (p. 37) of calling the semi-circular Klasma ‘Epiphonus’ and giving it musical sound.

^{page 129 note 2} Anthologia graeca Carminum Christianorum, Introd. cxxv.

^{page 129 note 3} Chrysanthus, Θεωρητικὸν μὲγα τῆς Μουσικῆς (Final edition, Trieste, 1832: reprint—which I use—Athens, 1911.) See especially, §§116–127.

^{page 129 note 4} Op. cit. c. 3; nor yet in the Codex Chrysander. Cf. c. 5.

^{page 130 note 1} Op. cit. § 129, where his list of Hypostases, excluding the Kr. is given.

^{page 130 note 2} A single Gorgon halves the unit of beat, making, e.g., two semiquavers; a two-fold Gorgon makes a triplet; a three-fold Gorgon makes four demi-semiquavers. (These latter are modern inventions.) Gastoué, op. cit. 3, cf. 43, 46, holds that the Gorgon in the MSS. had the same function as now.

^{page 130 note 3} Op. cit. 71.

^{page 131 note 1} Cf. Fleischer, op. cit. 64 and 72 (where his explanation of the use of the Gorgon is exactly the opposite of mine).

^{page 131 note 2} Chrysanthus, 133; Gastoué, op. cit. 36 (No. 32); Gaisser, op. cit. 9, etc.

^{page 131 note 3} P. 39, Gastoué (ibid.) gives no explanation; Gaisser and Thibaut, op. cit. do not even mention it.

^{page 131 note 4} P. 59.

^{page 132 note 5} Riemann, 40; Gaisser, l.c.

^{page 132 note 6} P. 73.

^{page 132 note 7} Cf. the use of Tromikon in the practice example reproduced by Fleischer, , Facs. p. 27Google Scholar; and Riemann's transcription, op. cit. 41. The majority of Hypostases or subsidiary signs are known to have been used in the manner above explained. (Tromikon occurs once in the four-fold Hallelujah reproduced in this article, p. 132.)

^{page 133 note 1} P. 71, Nos. 5 and 6; and exx. passim.

^{page 133 note 2} There is obviously some difference between such passages as and such fine shades of vocal effect may have been in the mediaeval composers' minds.

^{page 134 note 1} Cf. his transcriptions passim; e.g. No. xxx. pp. 38–9, where the effects of Diple and Double Apostrophus are repeatedly ignored.

^{page 134 note 2} The same thing may be easily seen if the Hallelujah above quoted be marked off according to Fleischer's rules. The 4-time breaks down completely.

^{page 134 note 3} Pp. 40–41.

^{page 135 note 1} Ibid. 41–43

^{page 136 note 1} Op. cit. Pl. VIII, and transcription, pp. 48–50.

^{page 136 note 2} I have pleasure in thanking the Librarian of Trinity College and also the sub-librarian (Mr. C. Hurry) whose kind help enabled me at a time of considerable difficulty to make a detailed study of the two valuable Byzantine Musical MSS. in the College Library.

^{page 136 note 3} E.g. in Ode i. l. 1 below, it would be unsuitable. Fleischer and others omit the subordinated signs entirely. My view agrees more or less with Riemann's here. If an ascending sign be annulled by a descending, the procedure would be the same.

^{page 136 note 4} Examples in Ode i. ll. 2, 3 below.

^{page 137 note 1} Cf. Ode i. l. 1 over καί.

^{page 137 note 2} The reason for this exception seems to be that the same pair of symbols (though with different meaning) was familiar in the Old or Linear Notation.

^{page 137 note 3} Cf. Ode i. l. 4.

^{page 142 note 1} Op. cit. Pl. V.

^{page 142 note 2} Op. cit. 52, 3.

^{page 143 note 1} These hymns begin on f. 102 b. The author is given in some MSS. as Andrew of Jerusalem.

^{page 143 note 2} Byz. Zeitsch. xx. p. 481.

^{page 143 note 3} Various forms of this mode are now in use. One called λέγετος begins from e (e.g. Christ, W. and Paranikas, , Anthologia, cxxxvii.).Google Scholar