lation into modern French

assonants.

Mirthful most when most they murmur.
And the envious Nymph of Air,
Seeing earth so richly studded
With the flowers of many springs,
Joined in this that is the youngest.
Has unto her azure plain,

Flowers of other kinds conducted;
Which, upborne on myriad wings,
Living nosegays float and flutter."*

Two years ago M. Petit de Julleville published a version of the entire "Chanson de M. Julleville's trans- Roland" in modern French verse, with assonant rhymes. It is a very remarkable achievement, and must have cost a world of labour. But French poetry has drifted so far from any of its forms in the period when assonants could please the ear, that it may be doubtful whether M. de Julleville's version will ever become popular, except with those who could enjoy the original; and whether the ordinary French reader would not prefer a version altogether unrhymed, like that of M. Gautier, M. Genin, or M. d'Avril. I give one of M. de Julleville's stanzas, that the reader may judge with what skill he has performed so difficult a task.

ORIGINAL.

Sansun li Dux vait ferir l'almacur,
L'escut li freinst k'est ad or e à flurs
Li bons osbercs ne li est guarant prud
Le coer li tranchet le feie e le pulmun
Que mort l'abat cui qu'en peist o cui nun
Dist l'Arcevesque cist colp est de Barun.

* Mr. Longfellow speaks of these translations in even warmer terms than Mr. Ticknor. [D. F. MacCarthy died on the Good Friday of 1882.]

M. de Julleville's version is as follows:

Et Samson 'frappe l'Emir; il brise en deux
Son riche ecu couvert d'or et de fleurs
Le bon haubert le garantit trop peu
Tranche le foie le poumon et le coeur

Et mort l'abat soit tant pis, soit tant mieux
Turpin s'ecrie ce coup est d'un vrai preux.'

66

I have to add that there is in the Oxford MS., at the end of each stanza or set of assonants,. The refrain the curious word or combination of letters, Aoi.

A O I, which has to this hour remained a puzzle to the critics; some consider it equivalent to the English "away," others regard it as a musical notation, others as a simple refrain or burthen.

tian MS.

In saying that the Bodleian manuscript is the only known copy of the "Roland," I mean the The Veneonly copy in its original form, the grammatical langue d'oil of the eleventh century. There is in truth, in the library of St. Mark, in Venice, a manuscript containing, with some variations, the whole of the "Song of Roland," down to the return of Charlemagne into France. At that point begins a total departure from the Oxford version, and a number of adventures are introduced, which are plainly later additions. The date of this manuscript M. Gautier assigns to the years 1230-40. The language is an Italianized French, such probably as the traveller, to his distraction, may still hear on the borders of Piedmont. The jongleurs, we may conceive, when

* In_Mr. Ticknor's letter to Mr. MacCarthy, which I have cited above, he says, speaking of assonant rhymes, "Would it not be amusing to have the experiment tried in French ?" Here is the experiment tried to the full.

they plied their vocation in places where the dialect was different from the language of the poem, altered the text viva voce, so as to make themselves understood by their hearers; and afterwards it was found convenient to have it written out in its altered form. The Venetian text has been of the utmost service to the editors of the Bodleian MS. It has thrown light on difficult passages, supplied valuable variants and the means of filling up lacunæ. An excellent edition of the Venetian manuscript was published in 1877, by Professor Eugen Kölbing. This version is in leashes of assonants like the Bodleian.

But as the French ear grew more cultivated, it became intolerant of the merely assonant rhyme; The rifac hence the rifaccimenti (remaniements), or cimenti. refashionings, of the poem according to the taste of a later time. The chief feature of these rifaccimenti is the change of the assonant into a complete rhyme; still preserving the leash or stanza. This was not to be done without much labour, nor without taking considerable liberties with the original. It was an operation not quite the same as changing blank verse into rhyme, because the assonants were occasionally perfect rhymes, but it was in great degree the same, much as if the Paradise Lost were laboriously turned into rhyme, by inferior artists. Having had to take so great a liberty with the original, all other liberties seemed little, and thus the poem became utterly defaced. M. Gautier mentions six manuscripts

*La Chanson de Roland. Genauer Abdruck der Venetianer Handschrift IV." Besorgt von Eugen Kölbing: Heilbronn, 1877.

of these rifaccimenti. One of them, of the fourteenth century, is in the library of Trinity College, Cambridge; another, known as the "Versailles MS.," is now in the Bibliothèque Nationale in Paris. This latter version M. Francisque Michel has appended, under the name of 66 Roncesvaux," to the second edition of his "Song of Roland," published in 1869. A comparison of the first stanza of the Bodleian MS., with the same stanza in the "Roncesvaux," will give us a clear idea of the nature of the remaniement.

CHANSON DE ROLAND.

Carles li reis nostre emperere magnes
Set anz tut pleins ad estet en Espaigne
Tresqu' en la mer cunquist la tere altaigne
N'i ad castel ki devant lui remaignet
Murs ne citet n'i est remés à fraindre
Fors Sarraguce kest en une muntaigne
Li reis Marsilies la tient ki Deu n'enaimet
Mahummet sert e Apollin reclaimet
Ne s poet guarden que mals ne li ateignet.

MS. OF VERSAILLES.

Challes li rois à la barbe grifaigne
Sis ans toz plens à este' en Espaigne
Conquist la terre jusqu à la mer alteigne
En meint estor fu ve'ue s'enseigne
Ne trove borc ne castel qu'il n'enplaigne
Ne mur tant aut qu'à la terre n'enfraigne
Forz Saragoze au chief d'une montaigne
La est Marsille qui la loi Deu n'en daigne
Mahommet sert, mot fait folle gaaigne
Ne poit durer que Challes ne le taigne
Car il n'a hom qu'à lui servir se faigne
Fors Guenelon que il tint por engeigne
Jamais n'ert jor que li rois ne s'en plaigne.

It may be seen in some slight degree from the above what liberties the refashioner permitted himself. Indeed, when once the mania of recasting the earlier

poems set in, the new versifier seems to have placed his chief glory in dragging in as many rhymes as possible into one stanza, and never knew how to leave off as long as he was able "to make it clink." It was in imitation of these models that Scott framed the commencement of the "Lay of the Bloody Vest," which Blondel sings to King Richard in "The Talisman."

History of the legend. Germany.

It

To return to the history of the poem. It must have had an early and wide success. was translated into Latin and then into German verse by a priest Conrad (Chuonrat), at the request of Duke Henry, whom M. Genin identifies with the Emperor Henry the Lion, and the request was made by desire of Henry's wife, Matilda, daughter of Henry II. of England. The date of this translation M. Genin fixes at from 1173 to 1177. In somewhat more than half a century, the "Ruolandes Liet" itself had to undergo a rifaccimento. The German language was then in such a state of flux and transition that even in that short space it had materially changed. The new adapter is known by the name of Stricker; and his poem was published under the title of "Karl." To follow the later German developments of the legend would be to transgress the limits of this Introduction.

Scandinavia.

The poem spread rapidly through the Scandinavian countries. The "Karlomagnus Saga " is an Icelandic compilation of the thirteenth The "Kar- century, consisting of translations from the French of all the current legends concerning Charlemagne; the eighth of these is

lomagnus Saga."