the separation from his wife. (See letter of October 9, 1821, Letters, 1901, v. 391.)
On his leaving England for the Continent, April 25, 1816, the fragment was left behind. Most probably the MS. fell into his sister's hands, for in October, 1821, it was not forthcoming when Byron gave directions that Hobhouse should search for it "amongst my papers." Ultimately it came into the possession of the late Mr. Murray, and is now printed for the first time in its entirety (vide post, pp. 453-466: selections were given in the Nineteenth Century, August, 1899). It should be borne in mind that this unprinted first act of Werner, which synchronizes with the Siege of Corinth and Parisina, was written when Byron was a member of the sub-committee of management of Drury Lane Theatre, and, as the numerous stage directions testify, with a view to stage-representation. The MS. is scored with corrections, and betrays an unusual elaboration, and, perhaps, some difficulty and hesitation in the choice of words and the construction of sentences. In the opening scene the situation is not caught and gripped, while the melancholy squalor of the original narrative is only too faithfully reproduced. The Werner of 1821, with all its shortcomings, is the production of a playwright. The Werner of 1815 is the attempt of a highly gifted amateur.
When Byron once more bethought himself of his old subject, he not only sent for the MS. of the first act, but desired Murray "to cut out Sophia Lee's" (vide post, p. 337) "German's Tale from the Canterbury Tales, and send it in a letter" (Letters, 1901, v. 390). He seems to have intended from the first to construct a drama out of the story, and, no doubt, to acknowledge the source of his inspiration. On the whole, he carried out his intention, taking places, characters, and incidents as he found them, but recasting the materials and turning prose into metre. But here and there, to save himself trouble, he "stole his brooms ready made," and, as he acknowledges in the Preface, "adopted even the language of the story." Act ii. sc. 2, lines 87-172; act iii. sc. 4; and act v. sc. 1, lines 94-479, are, more or less, faithful and exact reproductions of pp. 203-206, 228-232, and 252-271 of the novel (see Canterbury Tales, ed. 1832, vol. ii.). On the other hand, in the remaining three-fourths of the play, the language is not Miss Lee's, but Byron's, and the "conveyance" of incidents occasional and insignificant. Much, too, was imported into the play (e.g. almost the whole of the fourth act), of which there is neither hint nor suggestion in the story. Maginn's categorical statement (see "O'Doherty on Werner," Miscellanies, 1885, i. 189) that "here Lord Byron has invented
