last notable edition of the Poetics, that of Bywater (1909), the text occupies 45 pages out of 431. The Poetics makes about a hundredth part of the extant works of Aristotle.

Though never long, it doubtless once was longer; and it probably was associated with another work of Aristotle, now represented only by fragments, his dialogue On Poets, and with his Homeric Problems. In the same group of writings were the Peplos and the Didascalia, the latter a history or record of the Greek dramatic contests, with the names of victors and similar data. The dialogue On Poets seems to have been of a more literary character. At all events the Poetics does not now resemble the finished work of that Aristotle whose style was praised by Cicero and Quintilian. What we have of it may be the notes that served the master through various years for some part of his lectures; such notes he would expand in oral discussion, adding examples, reconciling apparent contradictions, solving difficulties with his pupils in the Peripatetic fashion. The defence of Homer toward the end of the book connects it with the Problems, and the last chapter, on the relative merits of epic and tragic poetry, has the look

of an embryo dialogue. Or we may have in the Poetics the notes of some person who attended Aristotle's lectures and colloquies-a hypothesis that would explain omissions and discrepancies that troubled the last generation of modern scholars. Or, finally, the work may be the grudging abstract by some student of the Alexandrian age or later, who took the essentials from one book or section by the master, and therewith joined what seemed important or germane from one or two others. There may have been several steps in the reduction of the Poetics to its present state, which it had reached some time before the sixth century A. D. If so, the history of the treatise would resemble that of Greek learning as a whole in the gradual decay of scholarship and science, until in the Dark Ages the rich and detailed investigations of the Alexandrian period-in biology, for example—had dwindled to the barest epitomes. There is at least one unexpected gap in our Poetics. The treatise does not fulfil its own promise regarding a discussion of comedy; certain scholars have found the promise redeemed in a fragment known as the Tractatus Coislinianus, the merest outline, descended from a body of

critical doctrine, now lost, of wider range than the extant Poetics.

We have no adequate knowledge about the composition of Aristotle's works in literary criticism. His Rhetoric, though the text is now corrupt, fared better in the Græco-Roman world than the Poetics, meeting the practical needs of Roman orators. We do not know when either treatise was written, yet there has been a tendency to regard the Poetics as the earlier. Doubtless both existed side by side during some part of Aristotle's activity as teacher, and underwent occasional revision at his hands. After his death the Rhetoric was included in a body of like treatises, now mostly of uncertain origin. But the Poetics is the only technical discussion of its subject that has come down to us from ancient Greece. For the study of Greek art, including poetic art, it is, after the masterpieces themselves, the most valuable document we have from antiquity.

In its own time, however, it was not a solitary work; and it had predecessors as well as contemporaries in its field. Here Aristotle did not, as in the Topica, feel that he labored as a pioneer, but had models on which to

improve. In Poetics 8 he suggests that Homer may have worked with conscious art. If a full-fledged theory of rhetoric came from the Sicilian to the Athenian orators, why, we may ask, should not a theory of the poetic art come to the Athenian drama, if not from Sicily, then from Asia Minor-from Miletus or Smyrnaalong with a body of epic tradition that furnished subject-matter for the Attic stage? The more we learn of early Ægean culture, and of its persistence at the Sicilian and Ionian fringes when the centre was swept away, the greater seems the debt of Athens to that culture for the seeds of art and science. Some notions of Homeric rules of art, accordingly, may have drifted down to the predecessors of Aristotle with the Iliad and the Odyssey. In the Republic Plato makes Socrates speak of the 'ancient quarrel between philosophy and poetry.' How early did the naughty Homeric tales of the gods, of Mars and Venus, become a topic of debate between moralist and literary critic? The Poetics notes that Xenophanes (fl. 530 B. C.) thought them very bad, and implicitly takes issue with the Republic for likewise condemning them with no appeal to standards of art.

But the foregoing are vaguer considerations. There is evidence in Plato's Phædrus, as in the Poetics, that Sophocles consciously observed dramatic laws. And, among other sayings, he declared that Eschylus 'did right without knowing why'; he himself, then, composed aright, knowing why. The Poetics records a maxim of Agathon on dramatic probability; and indeed, in acting and staging their plays, and in training the chorus and actors, the tragic poets must have reflected much on their art. That the great dramatists had a store of reflections is evinced by Aristophanes' Frogs, which in effect is the work of a great literary critic, and shows the poet to be familiar with tragic technique, and with stock terms and methods of criticism; his comic purpose should not blind us to his actual knowledge. Another (lost) play of Aristophanes was itself called Poesis, while of his contemporaries Plato (not the philosopher) produced a Poet and Poets, and Nicochares likewise a Poet. These comedies were the forerunners of those with similar titles in the age of Aristotle: Poets and Poetry by Alexis, Paesis by Antiphanes, and a Poet each by Biottus and Phoenicides.