significant. In Gottfried Baist’s argument, based on a rigid division between early lyric poetry (as Portuguese) and early prose (as Spanish), the Leonoreta lyric, far from being a stumbling-block, is actually a sign of the Spanish origin of _Amadis_: as a fragment (14th c.) of a prose _Tristan_ exists in Spanish, and five Portuguese Tristan _lais_ figure in the _Cancioneiro Colocci-Brancuti_, so the Leonoreta poem belongs to a Spanish _Amadis_ in prose. But although the priority and relations of early Portuguese and Spanish prose works are intricate and have not yet been thoroughly studied, it is clear that in many cases versions have been more carefully preserved in conservative Spain, while the Portuguese through neglect, fire, and earthquake have perished, and also that the natural tendency and development of prose, in view of the growing power of Castille and the greater pliancy of the Portuguese, was from Portuguese to Spanish, not from Spanish to Portuguese. And in one instance at least we have an early Portuguese prose work of the first importance, the _Demanda do Santo Graall_, which with its gallicisms can by no stretch of imagination be accounted a version from the Spanish. It is plainly legitimate to hold that the story of Amadis was first reduced to book form in the Peninsula in precisely the same way as was the story of Galahad, i.e. as a fourteenth-century Portuguese adaptation with the French text in view. Nicholas d’Herberay des Essarts, we know, claimed to have discovered fragments of _Amadis en langage picard_, Jorge Cardoso (1606-69) declared that Pero Lobeira translated _Amadis_ from the French,[139] and Bernardo Tasso, whose _Amadigi_ appeared in 1560, believed (_non è dubbio_) _Amadis_ to be derived _da qualche istoria di Bretagna_. Nor would the Portuguese, for all their familiarity with the story and topography of the Breton cycle, be likely to compose original works dealing with Vindilisora (Windsor) or Bristoya (Bristol). Unhappily, however deep may be our conviction (a conviction which stands in no need of antedating Hebrew versions of the 1508 _Amadis_) that the Peninsular _Amadis_ was originally Portuguese, it has now ceased to belong to Portuguese literature; another instance, if we may beg the question, of the gravitation to Spain. The Portuguese text, of which a copy, according to Leite Ferreira, existed in the library of the Duques de Aveiro in the sixteenth century (1598), and, according to the Conde da Ericeira, in the library of the Condes de Vimieiro in the seventeenth (1686), is still missing, as it was in 1726. FOOTNOTES: [104] Portuguese is then _uma lingua coherente, clara, um instrumento perfeito para a expressão do pensamento, cuja maior plasticidade dependerá apenas da cultura litteraria_, F. Adolpho Coelho, _A Lingua Portugueza_ (1881), p. 87. [105] See _supra_, p. 48. [106] See p. 160.