The word romancero in modern Spanish is more or less strictly applied to a special form of verse composition, a narrative poem written in lines of sixteen syllables which adhere to one single assonance throughout. Originally the term was applied to those dialects or languages which were the offspring of the Roman or Latin tongue--the spoken language of old Rome in its modernized forms. Later it came to imply only the written forms of those vernaculars, and lastly the poetic lyrico-narrative form alone, as above indicated. The romancero therefore differs from the romance in that it is written in verse, and it is plain from what has just been said that the name 'romance' was the product of the transition period when the term was intended to describe the written output of the more modern forms of Latin-Castilian, Portuguese, French, and Provençal, whether couched in prose or verse. We have seen that practically all the romances proper, as apart from the cantares de gesta--that is, such compositions as Amadis, Palmerin, and Partenopex--were written in prose. But the romancero was first and last a narrative in verse. Indeed, the three tales recounted in the last chapter are of the romancero type--a form, as we shall see, which gained quite as strong a hold upon the lower classes of the Peninsula as the romance proper did upon the affections of the hidalgo and the caballero. In a word, the romancero is the popular ballad of Spain. In a previous chapter I attempted to outline the several types of the Spanish ballad, or romancero, as follows: (1) Those of spontaneous popular origin and early date. (2) Those based upon passages in the chronicles or cantares de gesta. (3) Folk-ballads of a relatively late date. (4) Those later ballads which were the production of conscious art. We can thus class Spanish ballads more broadly into: (1) Those of popular origin. (2) Those which have their rise in literary sources. As regard class (1) of the first quaternion, like Sancho Panza I have no intention of indicating how old these may be. The fiercest controversy has raged round this question, but, as I have already indicated, it would be strange indeed if no vestiges of early Castilian folk-song had come down to us in an altered form. Folk-song, in my view, has as great a chance of survival as custom or legend, and we know how persistent these are in undisturbed areas, so I see no reason to doubt that a certain number of the original ballads of Spain have come down to us in such an altered form as would, perhaps, render them unrecognizable to their makers, just as the ancient Scottish romance of Thomas the Rhymer would not have been recognized in its later form by the singer who composed it.