Page:On translating Homer (1905).djvu/136

 modern music. But if the poet, at our request, instead of singing the verses, read or spoke them, then from the loss of well-marked time and the ascendency reassumed by the prose-accent, we should be as helplessly unable to hear any metre in them, as are the modern Greeks.

I expect that Mr Arnold will reply to this, that he reads and does not sing Homer, and yet he finds his verses to be melodious and not monotonous. To this, I retort, that he begins by wilfully pronouncing Greek falsely, according to the laws of Latin accent, and artificially assimilating the Homeric to the Virgilian line. Virgil has compromised between the ictus metricus and the prose accent, by exacting that the two coincide in the two last feet and generally forbidding it in the second and third foot. What is called the 'feminine cæsura' gives (in the Latin language) coincidence on the third foot. Our extreme familiarity with these laws of compromise enables us to anticipate recurring sounds and satisfies our ear. But the Greek prose accent, by reason of oxytons and paroxytons, and accent on the ante-*penultima in spite of a long penultima, totally resists all such compromise; and proves that particular form of melody, which our scholars enjoy in Homer, to be an unhistoric imitation of Virgil.

I am aware, there is a bold theory, whispered if not published, that,—so out