(See Vol. I. p. 128.)
As the earliest ballads were not by known authors, but were gathered at different times, from the traditions of the people, it is impossible to understand their history without understanding something of the history of the Ballad-books in which they are found. A sketch of such a history has been written, with much knowledge of the subject, by Ferdinand Wolf, and is found in the “Jahrbücher der Literatur” (Band CXIV., Wien, 1846, pp. 1-72). I do not willingly enter into a discussion so peculiarly within the province of this distinguished scholar; but, as I possess, or have seen, several very early Ballad-books which he does not mention, and am besides unable to agree with him as to which is the oldest of them all, and therefore the most important, I will, as briefly as I can, give my views of this obscure branch of bibliography; confining myself, where it is possible to do so, to what has not before been published, and touching the whole matter only so far as it concerns the history of Spanish poetry.
A considerable number of ballads, printed on one or more sheets, in black letter, for popular use, may still be found. Such are “El Conde Alarcos”; “El Moro Calaynos”; a collection of twelve separate pieces, and a collection of fifty-nine, sold at Heber’s sale; with others noticed by Brunet, under the head of Romances Séparées, in his article “Romanceros.” But they are all without dates; it is extremely uncertain when any one of them was printed; and it seems to me, judging from those I have seen, to be more probable, that they were taken from collections now known to exist or to have existed, than that they helped to make up those collections,—the oldest of which claims to have been taken from the memories of the people, and from imperfect manuscript copies circulating only for popular use.
1. The first separate collection of ballads ever published was, I think, the one printed at Saragossa, under the title of “Silva de Varios Romances,” by Stevan G. de Nagera, in two parts, 1550. (See Brunet, Manuel du Libraire, ed. 1843, art. Silva.) I have seen a copy of this Silva belonging, in 1838, to M. Henri Ternaux-Compans, of Paris. In a prefatory address to the First Part, the collector says, “I have taken the trouble in this Silva to bring together all the ballads that have come to my knowledge”; adding afterwards, “It may be that some, though very few, of the old ballads are wanting, which I have not inserted, either because they did not come to my knowledge, or because I did not find them so complete and perfect as I wished. Nor do I deny, that, in some of those here printed, there may be an occasional error; but this is to be imputed to the copies from which I took them, which were very corrupt, and to the weakness of memory of some persons, who dictated them to me, and who could not recollect them perfectly. I did all I could to obtain the least faulty that were to be had, and had no little trouble to collect and amend them, and add to some that were imperfect. However, I wished they should stand in some order, and so I placed, first, those of devotion and from the Holy Scriptures; next, those that relate Castilian stories; next, those of Troy; and, lastly, those that relate to affairs of love.” After these ballads, which fill one hundred and ninety-six leaves, he gives us twenty-five leaves of canciones, villancicos, and chistes, or jests, among which, at folio 199, is the well-known witty Dialogue of Castillejo and his Pen. At the end of the First Part, folio 221, we have the following Address to the Reader, in which the collector has evidently changed his mind about having obtained all but a “very few of the old ballads” known to exist; for he now says: “Some of my friends, as they knew I was printing this ‘Cancionero,’ brought me many ballads, in order that I might insert them; but as we were coming to the end of the printing, I chose not to put them in, since they would interrupt the order that had been begun; but rather to make another volume, which will be the Second Part of this ‘Silva de Varios Romances,’ which is now in the press. Vale.”
This “Segunda Parte” was published in the same year, 1550, and consists of two hundred and three leaves of ballads, nineteen leaves of chistes, and two leaves of contents, at the end of which the “Impresor” says: “I did not wish to put into this part any more of those short chistes, because, if God pleases, they will be put into the Third Part, with other things agreeable to the curious reader. Vale.” I know of no copy of this Third Part; but it is possible it was printed, because, in the “Silva de Varios Romances,” of which Wolf and Brunet mention several editions between 1578 and 1673, and of which I possess that of 1602, the title-page declares that it contains “los mejores romances de los tres libros de la Silva.”
2. The first two parts, however, combined into one, but omitting the chistes, etc., soon appeared at Antwerp, printed by Martin Nucio, a well-known publisher, with considerable additions, but without the date of its publication. The Preface is in nearly the same words with that of the Silva of Nagera, Parte I.; but, when it announces the arrangement of the ballads, it changes their order, and puts “first, those that speak of France and the Twelve Peers; then, those that relate Castilian stories; then, those of Troy; and, lastly, those that treat of affairs of love.” Some of the ballads of the Saragossa collection are omitted, and the whole is called “Cancionero de Romances.” There is a copy of it in the Bibliothèque de l’Arsénal at Paris; and that it is subsequent to the Saragossa Silva, and taken from it, seems certain, because one must be taken from the other, and the note at the end of the Silva, Parte I., shows that the Saragossa Silva was collected and printed at different times; while the arrangement of the ballads in the Cancionero of Antwerp shows that they were necessarily all present to the editor when he put his work together. Besides, how should Nucio collect ballads from the memories of the people around him at Antwerp, where there were few Spaniards, except soldiers? And how much less valuable would be any collection made there than one made in Spain?
3. Again, a “Cancionero de Romances” occurs, printed “En Envers en casa de Martin Nucio, MDL.,” a copy of which is in the Bibliothèque de l’Arsenal of Paris. It has the same Preface with the one last mentioned, from which it differs only in omitting seven of its ballads, and inserting thirty-seven others. The errors noted in the one without date, at folios 272. b, etc., are corrected in this one, dated 1550, and prove it to be the subsequent edition of the two,—a fact necessarily inferred, also, from the additions it contains.
4. This edition of 1550 seems to have been issued with different title-pages, for Wolf says there is a copy of it in the Imperial Library at Vienna, dated 1554. But nearly all the copies now known to exist bear the date of 1555, under which this collection is best known, and is commonly cited. It is absolutely the same work with the copy at the Bibliothèque de l’Arsenal, dated 1550, ballad for ballad, and page for page; and as there is no appearance that the title-page of the copy at the Arsenal has been tampered with, we are to suppose that three editions of the collection of ballads made at Saragossa in 1550 appeared in the course of that year; two of which were published by Martin Nucio, at Antwerp. That all three are only one work is apparent from the circumstance, that their ballads are generally the same, and that they have the same Preface, a little changed in the second and third editions to meet the changes in the ballads contained in them. They are all in 18mo. The first, taking both its parts together, fills four hundred and thirty-six leaves; the second, two hundred and seventy-six; and the third, three hundred. Several reprints of the last are given by Wolf; namely, Antwerp, 1568 and 1573; Lisbon, 1581; and Barcelona, 1587 and 1626.
Subsequent to the Silva of Saragossa, we have several collections of ballads, that are noticed in the text,—such as those of Sepulveda, 1551, Timoneda, 1573, Linares, 1573, Padilla, 1583, Maldonado, 1586, and Cueva, 1587,—consisting chiefly or entirely of ballads written by their respective authors. At last, an attempt was made to gather another Romancero from all the sources, whether of books, memory, or tradition, that were open to its collectors,—the true principle on which the popular Spanish Romanceros have always been compiled. It seems to have been begun at Valencia, when the first volume of the “Flor de Varios y Nuevos Romances, Primera y Segunda Parte,” collected by Andrés de Villalta, with a Third Part by Felipe Mey,—himself a poet and scholar as well as a printer,[458]—were printed in one volume, in 1593, though each of them had, probably, been printed earlier by itself. It is cited by Duran (Romances Caballarescos, Madrid, 1832, 12mo, Tom. I., Advertencia); and from the ballads he took out of it there can be no doubt, that its three parts differed little from the first three parts of the “Romancero General” printed somewhat later. The second volume of this collection, which is entitled “Quarta y Quinta Parte de Flor de Romances,” was collected by Sebastian Velez de Guevara, Racionero de la Colegial de Santander, and was printed at Burgos, in 1594, 18mo, one hundred and ninety-one leaves. It is apparently not the first edition, for the Aprobacion by Pedro de Padilla, and a permission to print it, are dated 1592, while the permission to print the present edition is dated August 11, 1594, and says it has been “otras veces impreso.” Probably the two parts were originally printed separately.
The third volume, and the most important, is entitled “Sexta Parte de Flor de Romances Nuevos, recopilados de muchos Autores, por Pedro de Flores, Librero,” and was printed at Toledo, in 1594, 18mo, one hundred and ninety leaves. It is the first edition, but the license seems to speak of a fourth and fifth part as if also made by Flores. In a ballad prefixed to this third volume, Flores is accused before Apollo of having taken great pains to collect its contents.