"There's just one bit I want to look through again," my father would say; but he would get carried away and rewrite the whole thing afresh. There were even occasions when, after posting the Proofs, my father would remember some particular words next day and correct them by telegraph. There, better than in a thousand generalizations, you see what the artistic conscience is. In a world in which authors, like solicitors, must live, it is, of course, seldom possible to take pains in this measure. Dostoevsky used to groan that his poverty left him no time or chance to write his best as Tolstoy and Turgenev could write theirs. But he at least laboured all that he could. Novel-writing has since his time become as painless as dentistry, and the result may be seen in a host of books that, while affecting to be fine literature, have no price except as merchandise. XXII.--THE THEORY OF POETRY Matthew Arnold once advised people who wanted to know what was good poetry not to trouble themselves with definitions of poetry, but to learn by heart passages, or even single lines, from the works of the great poets, and to apply these as touchstones. Certainly a book like Mr. Cowl's _Theory of Poetry in England_, which aims at giving us a representative selection of the theoretical things which were said in England about poetry between the time of Elizabeth and the time of Victoria, makes one wonder at the barrenness of men's thoughts about so fruitful a world as that of the poets. Mr. Cowl's book is not intended to be read as an anthology of fine things. Its value is not that of a book of golden thoughts. It is an ordered selection of documents chosen, not for their beauty, but simply for their use as milestones in the progress of English poetic theory. It is a work, not of literature, but of literary history; and students of literary history are under a deep debt of gratitude to the author for bringing together and arranging the documents of the subject in so convenient and lucid a form. The arrangement is under subjects, and chronological. There are forty-one pages on the theory of poetic creation, beginning with George Gascoigne and ending with Matthew Arnold. These are followed by a few pages of representative passages about poetry as an imitative art, the first of the authors quoted being Roger Ascham and the last F.W.H. Myers. The hook is divided into twelve sections of this kind, some of which have a tendency to overlap. Thus, in addition to the section on poetry as an imitative art, we have a section on imitation of nature, another on external nature, and another on imitation. Imitation, in the last of these, it is true, means for the most part imitation of the ancients, as in the sentence in which Thomas Rymer urged the seventeenth-century dramatists to imitate Attic tragedy even to the point of introducing the chorus.