more, and Sir Walter Scott, may well be excused the general censure. The former, living in and pandering to an age which invented and applied those delightful literary adjectives 'elegant' and ingenious,' may be pardoned with the more sincerity if one recalls the influence exercised on English letters by his publication. The latter, who played the part of Percy in the matter of Scottish ballads, and was nourished from his boyhood on the Reliques, printed for the first time many ballads which still are the best of their class, and was gifted with consummate skill and taste. Both, moreover, did their work scientifically, according to their lights; and both have left at least some of their originals behind them. There is, perhaps, one more exception to the general condemnation. Of William Allingham's Ballad Book, as truly a vade mecum as Palgrave's lyrical anthology in the same Golden Treasury' series, I would speak, perhaps only for sentimental reasons, always with respect, admiring the results of his editing while looking askance at the method, for he mixed his ingredients and left no recipe.

6

But in the majority of cases there is no obvious excuse for this omnium gatherum' process. The self-imposed function of most ballad editors appears to have been the com

pilation of rifacimenti in accordance with their private ideas of what a ballad should be. And that such a state of things was permissible is doubtless an indication of the then prevalent attitude of half-interested tolerance assumed towards these memorials of antiquity.

To-day, however, the ballad editor is confronted with the results of the labours, still unfinished, of a comparatively recent school in literary science. These have lately culminated in The English and Scottish Popular Ballads, edited by the late Professor Francis James Child of Harvard University. This work, in five large volumes, issued in ten parts at intervals from 1882 to 1898, and left by the editor at his death complete but for the Introduction-valde deflendus -gives in full all known variants of the three hundred and five ballads adjudged by its editor to be genuinely 'popular,' with an essay, prefixed to each ballad, on its history, origin, folklore, etc., and notes, glossary, bibliographies, appendices, etc.; exhibiting as a whole unrivalled special knowledge, great scholarly intuition, and years of patient research, aided by correspondents, students, and transcribers in all parts of the world, Lacking Professor Child's Introduction, we cannot exactly tell what his definition of a 'popular' ballad was, or what qualities in a ballad implied

exclusion from his collection-e.g. he does not admit The Children in the Wood: otherwise one can find in this monumental work the whole history and all the versions of nearly all the ballads.

It will be obvious that Professor Child's academic method is suited rather to the scholar than the general reader. As a rule, one text of each ballad is all that is required, which must therefore be chosen-but by what rules? Το the scholar, it usually happens that the most ancient and least handled text is the most interesting; but these are too frequently incomplete and unintelligible. The literary dilettante may prefer tasteful decorations by a Percy or a Scott; doubtless Buchan has some admirers: but the student abhors this painting of the lily.

Therefore I have compromised-always a dangerous practice and I have sought to give, to the best of my judgement, that authorised text of each ballad which tells in the best manner the completest form of the story or plot. I have been forced to make certain exceptions, but for all departures from the above rule I have given reasons which, I trust, will be found to justify the procedure; and in all cases the sources of each text or part of the text are indicated.

I am quite aware that it may fairly be asked:

Why not assume the immemorial privilege of a ballad editor, and concoct a text for yourself? Why, when any text of a ballad is, as you admit, merely a representative of parallel and similar traditional versions, should you not compile from those other variants a text which should combine the excellences of each, and give us the cream?

There are several objections to this course. However incompetent, I should not shrink from the labour involved; nor do I entirely approve the growing demand for German minuteness and exactitude in editors. But, firstly, the ballad should be subject to variation only while it is in oral circulation. Secondly, editorial garnishing has been overdone already, and my unwillingness to adopt that method is caused as much by the failure of the majority of editors as by the success of the few. Lastly, chacun a son goût; there is a kind of literary selfishness in emending and patching to suit one's private taste, and, if any one wishes to do so, he will be most pleased with the result if he does it for himself.

This lengthy apologia is necessitated by a departure from the usual custom of balladediting. For the rest, my indebtedness to the work of Professor Child will be obvious through

out.

Many of his most interesting texts were

printed for the first time from manuscripts in private hands. These I have not sought to

and care.

collate, which would, indeed, insult his accuracy But in the case of texts from the Percy Folio, where the labour is rather to decipher than to transcribe accurately, I have resorted not only to the reprint of Hales and Furnivall, but to the Folio itself. The whimsical spelling of this мs. pleases me as often as it irritates, and I have ventured in certain ballads, e.g. Glasgerion, to modernise it, and in others, e.g. Old Robin of Portingale, to retain it literatim : in either case I have reduced to uniformity the orthography of the proper names. Transcripts from other мss. are reproduced as they stand.

In the general Introduction I have tried to sketch the genesis and history of the ballad impartially in its several aspects, not for scholars and connoisseurs, but for those ready to learn. To supply deficiencies, I have added a list of books useful to the student of English ballads— to go no further afield. Each ballad also is prefaced with an introduction setting forth, besides the source of the text, as succinctly as is consistent with accuracy, the derivation, when known, of the story; the plot of similar foreign ballads; and points of interest in folklore, history, or criticism attached to the particular