tu, Melpomene.”  For this I took a different metre, which happens to be identical with that of a solitary Ode in the Second Book, “Non ebur neque aureum,” being guided still by my feeling about the individual Ode, not by any more general considerations.  I did not attempt a third until I had proceeded sufficiently far in my undertaking to see that I should probably continue to the end.  Then I had to consider the question of a uniform metre to answer to the Latin.  Both of those which I had already tried were rendered impracticable by a double rhyme, which, however manageable in one or two Odes, is unmanageable, as I have before intimated, in the case of a large number.  The former of the two measures, divested of the double rhyme, would, I think, lose most of its attractiveness; the latter suffers much less from the privation:  the latter accordingly I chose.  The trochaic character of the first line seems to me to give it an advantage over any metre composed of pure iambics, if it were only that it discriminates it from those alternate ten-syllable and eight-syllable iambics into which it would be natural to render many of the Epodes.  At the same time, it did not appear worth while to rewrite the two Odes already translated, merely for the sake of uniformity, as the principle of correspondence to the Latin, the alternation of longer and shorter lines, is really the same in all three cases.  Nay, so tentative has been my treatment of the whole matter, that I have even translated one Ode, the third of Book I, into successive rather than into alternate rhymes, so that readers may judge of the comparative effect of the two varieties.  After this confession of irregularity, I need scarcely mention that on coming to the Ode which had suggested the metre in its unmutilated state, I translated it into the mutilated form, not caring either to encounter the inconvenience of the double rhymes, or to make confusion worse confounded by giving it, what it has in the Latin, a separate form of its own.

The remaining metres may be dismissed in a very few words.  As a general rule, I have avoided couplets of any sort, and chosen some kind of stanza.  As a German critic has pointed out, all the Odes of Horace, with one doubtful exception, may be reduced to quatrains; and though this peculiarity does not, so far as we can see, affect the character of any of the Horatian metres (except, of course, those that are written in stanzas), or influence the structure of the Latin, it must be considered as a happy circumstance for those who wish to render Horace into English.  In respect of restraint, indeed, the English couplet may sometimes be less inconvenient than the quatrain, as it is, on the whole, easier to run couplet into couplet than to run quatrain into quatrain; but the couplet seems hardly suitable for an English lyrical poem of any length, the very notion of lyrical poetry apparently involving a complexity which can only be represented by rhymes recurring at intervals.