qui lux es et dies", "Aurora lucis rutilat", "O Rex aeterne Domine", and "Mediae noctis tempus est". In the first of these hymns, which is probably the oldest, (having been persistently, though of course erroneously, ascribed to Ambrose himself), there are no polysyllables in which the regular accent does not coincide with the normal metrical stress. In the last named of the six, which is probably also the latest,(1) there are 8. In all the hymns together there are 18 such polysyllables, in a total of 236 lines. A decided majority of these (13) occur in the first halves of their respective lines, so that that those lines, if read accentually, present only the familiar initial versions, single or double. E. g. in the second of the hymns named we have It looks very much as if these polysyllables, where their accent affects only the first foot, or the first two feet, were meant to be read as in prose,-otherwise (perhaps) to be forced into the regular rhythmical scheme. If this can be believed, it would seem that at a very early date,- -as soon as the accentual principle was fairly grasped, the propriety and beauty of occasional inversions came to be appreciated. We should then explain the verses of Fortunatus somewhat differently: for his irregularities could be called ordinary inversions. His (1) See March's note: he assigns it to the 6th century. I should think, from the irregularities in the numbers of syllables (11. 33, 34) and the intermixture of trochaic lines (11. 46, 47) that it must be as late as that. It could hardly have been written before the accentual idea was firmly grasped. hymn would then be regarded not as written in quantity, with careful observance of prose accents, but as written accentually, with all convenient regard for the rules of quantity. This, indeed, seems the more plausible view, but in the present state of our knowledge it is hard to see how any certain inferences can be drawn. CHAPTER IV. Early Church Music. Syllabic verse. § 28. Introduction. The opinion advanced in the foregoing pages as to the origin and nature of the late Latin rhythmical versification is of course too obvious to be wholly novel,(1) but lately, in the search for more remote explanations, its obviousness has become obscured. The purpose of this chapter will be to explain away one of the most serious misunderstandings by which students have been led away from the plain path. There are in existence several specimens of late Latin verse of a strictly syllabic character, in which even the most strenuous upholder of the accentual theory can find no trace of the accentual principle. These have been regarded as proving (by obvious analogy) the absence of that principle from other forms of verse:(2)-but it is believed that they can easily be explained as not inconsistent with the accentual theory. Most of the early "rhythmical" verse was meant to be sung, and sufficient attention has not been paid-in the investigation of the "rhythms" themselves-to the exigencies of the music which accompained them. It is believed that a consideration of the diverse styles of music in use in the early churches will materially aid the present investigation. (1) See, for example, Gautier, I. 281 et seq., where similar conclusions are reached, though by somewhat questionable reasoning. (2) Cf. Meyer's treatment of Augustine's Psalm. Our information about early Christian music is scanty. "However persevering" says Fétis(1) "may be the historian's efforts to collect information about the music of the Roman church during the first four centuries, they must result only in a conviction of their futility, for no trustworthy document on that subject is now extant. A few words of Tertullian, dealing in generalities too vague to be of any use to us, are all that we can cite". It would seem, therefore, that most of what we can learn about the music of this period must be by inference from what we know of the state of music before the Christian era and of the state of music after the fourth century: but we shall find that there is some contemporary evidence which, though indirect, will be of no mean value. § 29. Greek music. Fortunately the aspect of the subject that is most nubilous is one which does not concern our present inquiry at all,-namely, the question as to the modes and keys employed in the early churches. The various modes in vogue at different times seem to have had none but the most indirect effect upon the time and rhythm of the melodies. These latter are the elements with which versification is most intimately connected, and they can be studied independently, so far as they need now be studied at all. The most striking characteristic of the religious melodies composed by the Greeks immediately before our era was this, that they followed closely, in time, the quantities of the syllables to which they were sung.(2) A long note was not sung to a short syllable, nor vice versa:-and in general each syllable was represented by a single note in the melody. A few measures from one (1) Vol. 4, p. 147. Tertullian's valueless testimony is given on p. 5 of the same volume. (2) The contrary view was formerly held, on the authority of a passage in Dionysius of Halicarnassus :-See Chappell, I. p. 172. But cf. also Monro, p. 117. of the musical inscriptions recently exhumed at Delphi will make this clear. (1) ἵ-να Φοῖβον φ' - δαῖ - σι μέλ-ψη-τε χρυ-σε-ο-κό-μαν. Here the two circumflexed syllables take two notes each. This is usually the case throughout the hymn, and in a few instances long syllables that are not circumflexed are treated similarly; but there is a general correspondence of note for syllable, and the musical time and rhythm agree always with the prosody. That this was universally the case admits no doubt. Changes in time may have been common,(2)-changes in tempo perhaps still more so,-but words and music would of course be affected by such changes exactly alike. The fragment given above dates probably from the 3rd century, B. C.(3) The other musical inscriptions discovered with it at Delphi are much shorter and more imperfect. They appear to be of much later date, the most considerable of them belonging apparently to the 1st century B. C. They present the same characteristics as the one just discussed, except that in the first place the time-values of the notes in one fragment undergo the changes necessary to fit them to a logaœdic measure, and in the second place the custom of suiting each syllable with a single note appears to have become a rule without exceptions. Even the circumflexed syllables have but one note apiece. § 30. The music of Ambrose. It is to be wished that authentic specimens might be given of Roman music (1) I copy part of the transcription given by Monro, p. 136. (3) Monro, p. 134. |