The teaching of _poetica_, from of old a part of _grammatica_, included extensive practise in Latin verse. This had early to take account of that dominance of stress which had gradually supplanted the ancient control by time. The characteristic medieval achievements in Latin lyric are the hymns. Radiating into other songs, even into humorous and satirical verse, the hymns were the common lyric fund of medieval Latin. As early as St. Ambrose they had created a new Latin poetry; and the beauty of their various art was not exhausted with Adam of St. Victor. Meantime they opened to the vernaculars those poetic possibilities of stanza which arise from the development of rime. Medieval poetic theory, on the contrary, went but a little way. Mainly pedagogical formulation, it lagged far behind the most characteristic medieval poetic advance, which was in verse narrative. Here is a sharp contrast with the Renaissance. The fifteenth century opens a long series of critical inquiries into poetic. The middle age, merging poetic with rhetoric in the schoolroom, was little concerned to make it tally with vernacular achievement. With the death in 1400 of Chaucer, whose criticism exposed this lack, the poetic of medieval narrative reaches its term. I owe to the unstinted courtesy and scholarly interest of a trustee of Barnard College, Mr. George A. Plimpton, the privilege of studying at leisure his manuscript of one of the most important Bolognese _dictamina_, the thirteenth-century _Candelabrum_. Far better than Boncompagno or Thomas of Capua, better even than Conrad, this unprinted manual exhibits _dictamen_ in both scope and method. My other debts are too manifold to rehearse. The bibliographical notes, if they recorded the reading of years, would defeat their proper object of serving further study. Therefore they have been made, as in the former volume, at once specific and strictly selective, applied to each chapter separately, and further indicated both in the index and on a page of recurring abbreviations after the table of contents. As I record gratefully my obligation for generous help with the proofs to my colleagues Professors Ayres, Clark, Krapp, McCrea, Moore, Perry, and Van Hook, and to my old friend, the Jesuit scholar Dr. Donnelly, I see further in such coöperation great promise for the progress of medieval studies. C. S. B. BARNARD COLLEGE COLUMBIA UNIVERSITY JANUARY, 1928. CONTENTS CHAPTER PAGE