John S. Tuckey first demonstrated the fact in 1963 in an admirable monograph in which he dated the composition of the manuscripts;1 this publication of the texts themselves offers additional proof. Thus, half a century after a spurious version was delivered to an unsuspecting public in the form of a children's Christmas gift book, the manuscripts are presented here for the first time as they came from their author's hand.

Paine was able to publish the "final complete work"-he said in 1923-because he turned up its essential last chapter in a great batch of unfinished stories and fragments several years after Clemens died in 1910.2 On the basis of incomplete evidence and wrong dating of manuscripts, Paine's successor as literary editor, Bernard DeVoto, argued that in completing The Mysterious Stranger, Mark Twain "came back from the edge of insanity, and found as much peace as any man may find in his last years, and brought his talent into fruition and made it whole again."' Two generations of readers have found the published tale as moving as DeVoto did. Although a very few readers and critics, notably Frederick A. G. Cowper and Edwin S. Fussell,' have been troubled by inconsistencies, especially in the final chapter, most have agreed that the melancholy fable, Twain's last important fiction, formed a kind of Nunc Dimittis.

The truth is that Mark Twain attempted at least four versions of the story, which survive in three manuscripts. The Mysterious Stranger represents, partially, the first manuscript in order of composition rather than the last, as DeVoto thought. None of the three is a finished work, although Twain did draft a "Conclusion of the book" for the third manuscript with the intent-never fulfilled-of completing this last version. Further, it is now clear that Paine, aided by Duneka, cut and bowdlerized the first manuscript heavily. He borrowed the character of the astrologer from the third manuscript and attributed to the new figure the grosser acts and speeches of a priest. Then he grafted the final chapter of the third manuscript to the broken-off first manuscript version by cutting half a chapter, composing a paragraph of bridgework, and altering characters' names. Speaking of his great discovery among the confusion of papers, Paine said, "Happily, it was the ending of the story in its first form." ° Although Paine's loyalty to Mark Twain was great and his rich accumulation of data about Mark Twain's life in Mark Twain: A Biography will always be valuable, two facts must be recorded here. Ile altered the manuscript of the book in a fashion that almost certainly would have enraged Clemens, and he concealed his tampering and his grafting-on of the last chapter, presumably to create the illusion that Twain had completed the story, but never published it. One bit of evidence proves this conclusively: in the all-important final chapter, on the manuscript the names "August" and "44," which Twain had given characters in the last version, are canceled, and "Theodor" and "Satan," characters in the first version, are substituted in Paine's hand.

A case can be made for Paine. When he and Duneka lifted the magician from the third manuscript, developed this figure into the astrologer, and used him as a kind of scapegoat, they thought they were acting to sustain and add to Mark Twain's reputation. They cut passages that they believed would offend Catholics, Presbyterians, and others for the same reason, and in cutting they did eliminate burlesque passages that clog the story. Moreover, as the experience of thousands of readers attests, the last chapter, although it was written for another version, does fit this version remarkably well. Certain "dream-marks" do suggest a dream-conclusion. But the major and inescapable charge in the indictment of Paine as editor of The Mysterious Stranger stands-he secretly tried to fill Mark Twain's shoes, and he tampered with the faith of Mark Twain's readers.

It follows that the serial text in Harper's Monthly Magazine (May through November 1916) and the text of the book (published in late October) possess no authority in the preparation of this edition. The text of the first edition remains chiefly an exhibition of the self-confident taste of the editor and his associate, Duneka-and, it seems likely, of their desire to get out another book by "Mark Twain." One depressing aspect of their misrepre-sentational editorial work is that they commissioned N. C. Wyeth, a well-known illustrator of children's books, to illustrate their altered text, and they let the designer place a fine color engraving of that nonentity, the "borrowed" astrologer, on the front cover.

The Order of the Manuscripts

Three of Twain's holograph manuscripts in the Mark Twain Papers at the University of California in Berkeley provide the copy-text of this edition. Typescripts of the first and third manuscripts, with a few authorial corrections, possess subsidiary authority. Mark Twain's titles for each, in the order of composition, were "The Chronicle of Young Satan," "Schoolhouse Hill," and "No. 44, The Mysterious Stranger." His manuscript workingnotes for the three versions, a long notebook entry about "little Satan, jr.," and a single discarded page of manuscript surviving from revision are included entire in appendixes. Explanatory notes follow.'The Textual Apparatus describes the texts, sets forth the editorial principles observed, lists the recovered cancellations, and gives all editorial choices or emendations. Here, as elsewhere in the University of California Press edition of The Mark Twain Papers, the intention is to set forth all the evidence for the making of the text.

The present dating of these works follows closely the conclusions of Tuckey, who in Mark Twain and Little Satan made a thorough examination not only of the manuscripts but of the whole body of documents in the Mark Twain Papers from 1897 through 1908, comparing papers, inks, and handwriting for dating clues and making skilled use of internal evidence as well.' Other literary evidence to be cited supports Tuckey's dating at every point.

Four versions of the narrative are to be distinguished in the three manuscripts:

Version A. The first may be called the "St. Petersburg Fragment" (Tuckey's "Pre-Eseldorf" pages).