Nor could anything else possibly happen, considering the methods which Boileau chose to adopt. [Sidenote: _Faults of method._] These methods are, First, the construction of a Horatian-Aristotelian bed to which everything has to be adjusted in point of principle; and of this it is not necessary to say more. Secondly, the suggestion as models, at every turn, of Latin and (much less often) Greek poets, utterly regardless of the change of circumstances. Thirdly, the method of criticism by Kinds—of laying down the rules for, and discussing the ends of, the abstract Pastoral, the abstract Elegy, the abstract Sonnet, Ode, and so forth, with only a few perfunctory eulogies of actual examples. In regard to the first and second it is enough to quote the critic’s own words against him, and ask him how Ovid or Tibullus can be a sufficient, can be even a safe, guide to a French love-poet, and how the marvels of ancient mythology can become the “machinery” of a modern epic? In regard to the third, the old battering-ram must be once more applied. Pastoral, Elegy, Ode, and the rest (except “Sonnet,” which is a _form_) are not unequivocal names applied to abstractly existing things, but mere tickets. Give me a poem, and I will tell you whether I think it a good or a bad poem, and why. You may, after that, if you have the time and care to take the trouble, classify it as epic or elegy, epigram or ode. But the box in which you choose to deposit it does not really matter in the least; and if it should so happen that there is no box ready, you must either make, or do without, one. Is the poem good or bad as poetry?—that is the _articulus stantis vel cadentis criticismi_.