purpose of acquiring the Arabic. This curious passage is cited by Du Cange, whose words I shall quote at length.

"Quòd vero suprà laudatus Scriptor ano nymus de Galliæ nostræ in Lingua Latina bar barie ante Caroli M. tempora, idem de Hispania post Saracenorum irruptionem testatur Alvarus: ubi neglectis et posthabitis Scrip turis Sanctis, earumque sacris interpretibus, quotquot supererant Christiani, Arabum Chaldæorumque libris evolvendis incumbebant, gentilitia eruditione præclari, Arabico eloquio sublimati, Ecclesiasticam pulchritudinem ignorantes, et Ecclesiæ flumina de Paradiso manantia, quasi vilissima contemnentes, legem suam nesciebant, et linguam propriam non advertebant Latini, ita ut ex omni Christi Collegio vix inveniretur unus in milleno hominum genere, qui salutatorias fratri posset rationaliter dirigere literas, CUM REPERIRENTUR ABSQUE NU

MERO MULTIPLICES TURBÆ, QUI ERUDITE CHALDAICAS VERBORUM EXPLICARENT POMPAS. Quod quidem abunde firmat; quæ de Elepanto Toletano suprà diximus. et inde satis arguimus unde tot voces Arabicæ in Hispanam, subinde sese intulerunt '."

Sed

We have then a complete refutation of Ritson's strongest objection; and perhaps had not the spleen of the writer been more powerful than the good sense and feeling of the man, he never would have hazarded the remark. And if judicial astrology, medicine, and chemistry, were of Arabian origin, and introduced into Europe a century at least before the crusades: if Pope Gerbert, or Sylvester II. who died A.D. 1003, brought the Arabic numerals into France, it is surely reasonable to suppose that these sciences, so

ན

1 Du Cange; Gloss. Med. Inf. Lat. Tom. I. Præfatio, p. xxxii. $31.

intimately connected with magical operations (and with fictions diverging from them) as to confer upon the possessor a title to supernatural agency, would extend their influence to the legendary stories, as well as to the manners of the west, which these very stories are admitted to describe! Yet, after all, it is not to be imagined that the introduction of eastern invention happened at one time, or in one age; it was rather the growth of many times, and of many ages-continually, though gradually augmenting, till it attained maturity.

The next hypothesis gives Armorica, or Bretany, as the source of romantic fiction. But to this, the same objections arise that have been started with respect to the rest. Mr. Ellis, in the introduction to his " Specimens of Early English Romances,” plausibly suggests that all are compatible. He

imagines" that the scenes and characters of our romantic histories were very generally, though not exclusively, derived from the Bretons, or from the Welsh of this island; that much of the colouring, and perhaps some particular adventures, may be of Scandinavian origin, and that occasional episodes, together with part of the machinery, may have been borrowed from the Arabians." Which is as much as to say, that each nation contributed something, and very likely they did; but which furnished the greater part, or which originated the whole, is just as obscure as before a "reconciliation" of opinions was projected. This conciliatory system will remind the reader of Boccacio's tale of The Three Rings, "the question of which is yet remaining."

Another supposition attributes the chief source of romantic fiction to classical and

1 Vol. I. p. 35.

mythological authors; that is, to the stories of Greece and Rome, somewhat altered by modern usages. To this belief Mr. Southey' and Mr. Dunlop seem to incline. The latter adds, that “after all, a great proportion of the wonders of romance must be attributed to the imagination of the authors." But when these wonders, similarly constructed, pervade the most remote countries, there must be something more than an author's imagination brought into the account. Consideration, however, is due to the idea of a classical origin; and this, blended with the rest, may help to make up a perfect system. Before I proceed to the attempt, I would advert to certain observations which Mr. Dunlop has promulgated in his "History of Fiction." He says, "It cannot be denied, and indeed has been acknowledged by Mr. Warton, that

1 Introduction to Amadis of Gaul.