[begin page 664]

An individual textual apparatus for each sketch provides everything needed to reconstruct the copy-text and Mark Twain's revisions (whenever any survives). Each apparatus usually, but not invariably, includes the elements described below. A description of texts follows the textual introduction in this volume; it identifies textually significant editions of Mark Twain's sketches and specifies the copies collated and examined in the preparation of this collection. A list of word divisions in this volume, which records ambiguous compounds hyphenated in the present edition at the end of a line, is given at the end of the entire apparatus to facilitate accurate quotation of Mark Twain's texts.

Textual Commentary

This section gives the copy-text and specifies the copy or copies used; it discusses problems or unusual features of the text; and under a subheading, “Reprintings and Revisions,” gives the history of the sketch and characterizes Mark Twain's revisions of it.

Textual Notes

This section discusses emendations or decisions not to emend: it calls attention to possible errors left unemended in the text, to problems in establishing particular readings, and to variants in the reprinting history which are especially problematic.

Emendations of the Copy-Text

This section records every departure in this edition from the copy-text, with the exception of the typographical features discussed above. It also records the resolution of doubtful or ambiguous readings. In each entry, the reading of this edition is given first, with its source identified by a symbol in parentheses; it is separated by a centered dot from the rejected copy-text reading on the right, thus:

[begin page 665]

A wavy dash (˜) to the right of the dot stands for the word on the left and signals that a mark of punctuation is being emended; a caret (‸) indicates the absence of a punctuation mark; and the symbol I-C follows any emendation whose source is not an authoritative text, even if the same correction was made in a derivative edition, thus:

Emendations marked with an asterisk in the left margin are discussed in the textual notes. A vertical rule indicates the end of a line in the copy-text, thus:

Italicized words in square brackets, such as [not in], [no ¶], and [torn], are editorial. Doubtful readings are recorded with the following notation:

Emendations and Adopted Readings

This section replaces Emendations of the Copy-Text when the text is established from radiating or composite texts. It records all variants, substantive and accidental, among the relevant texts, which are identified by abbreviations with superscript numbers; the numbers are assigned according to the chronology of publication and do not indicate relative authority of [begin page 666] the texts. Thus the following entry shows that three texts agree with each other against a fourth, and the majority reading has been adopted in this collection:

And the following entry shows that a compound hyphenated at the end of a line in one text and rendered solid in another is resolved in accord with the two that render it hyphenated:

An entry that rejects all of the radiating texts in favor of an editorial emendation appears as follows:

Diagram of Transmission and Historical Collation

These elements appear in the textual apparatus only for sketches reprinted or revised by Mark Twain. We give a diagram every time there is a chain of transmission, and it is essential for reading the entries in the historical collation. A list of the texts collated for each sketch immediately precedes the collation, which records all substantive variants in them. In addition, because Mark Twain is known to have concerned himself with revising emphasis (italics and exclamation points) as well as paragraphing, such variants in accidentals are likewise recorded. When Mark Twain demonstrably corrected or revised other accidentals—spelling, punctuation, and so on—in any of the surviving marked copies (YSMT, JF1MT, HWaMT, HWbMT, and MTSkMT), the full history of the particular accidental variant is also recorded.

In each collation entry the reading of this edition is given first, followed by symbols for the texts that agree with it; it is separated by a centered dot from its variants, which are identified by the appropriate symbols (given in the list of texts collated). A sample chain of transmission is given in figure 29 to facilitate understanding the examples that follow. Each transmission diagram in the individual apparatuses is essential to reading the collation for that sketch, because although the pattern of transmission is similar for many sketches, it varies in significant ways from sketch to sketch.

[begin page 667] [begin page 668]

Symbols joined by a dash (–) indicate that the reading appeared in the first text noted and was transmitted as far as the second. Thus the entry

indicates that the original dialect word was accurately transmitted from the Golden Era (GE) through JF1 and JF2 to JF4; but that Hotten altered it to “get” in his JF3, from where it was transmitted to HWa, then to HWb, and ultimately to SkNO, without being corrected by Mark Twain. A plus sign (+) indicates that the reading appears in the given text and in all printings derived from it. Thus the entry

indicates that a paragraph break appears at this point in GE, but not in JF1 or in any subsequent reprinting deriving from it. A more complicated entry, involving Mark Twain's revision, appears thus:

Here several complexities are recorded. The reading of the present edition and of the copy-text was, in this case, successfully transmitted from GE through JF1, JF2, and JF4 to MTSk; it was also transmitted from GE through JF1, JF2, and JF3 to HWa. Mark Twain encountered it there and revised it to the variant reading in HWaMT, which was incorporated in HWb and subsequently reprinted in SkNO. A still more complicated entry appears thus:

The original reading was transmitted successfully from GE through JF1 to JF2; it was also successfully transmitted from GE through JF1, JF2, and JF3 to HWa. JF4 altered “ate” to “eat” without authority. Mark Twain revised the JF4 text by striking out the passage following “and” and substituting an expressive long dash in MTSk. But when he revised HWaMT, he revised differently: he again deleted the matter after “and,” but substituted another dash and the phrase “here [begin page 669] is food for the imagination.” This revision was incorporated in HWb and subsequently reprinted in SkNO.

The historical collation preserves all substantive variants in the texts collated, whether or not they originated with Mark Twain. But the history of reprinting and revision given in the textual introduction permits us to make certain discriminations between variants which Mark Twain certainly made, and those which he could not possibly have made.

(1) Variants that first appear in JF2, JF3, JF4, Scrs, EOps, PJks, and HWa cannot be authorial.

(2) Variants that first appear in JF1, CD, MTSk, HWb, Sk#1, and SkNO may be authorial, but may also have been introduced by editors or compositors.

(3) Variants for which we have documentary evidence in the form of Mark Twain's autograph changes—YSMT, JF1MT, HWaMT, HWbMT, and MTSkMT (and a few stray examples of printer's copy in other forms)—are certainly authorial, and are so indicated in the collation by the symbol MT. Any variant that arises in JF1, HWb, or SkNO can be certainly attributed to Mark Twain when the revision appears first in the marked printer's copy.

The textual apparatus may contain other elements that are more or less self-explanatory. In the items where copy-text is a holograph, we include a section called Alterations in the Manuscript, which reports the author's cancellations, substitutions, and revisions. Essential corrections that Mark Twain made as he wrote or reread his work are not recorded: letters or words that have been mended or traced over, or canceled and rewritten merely for clarity; false starts and slips of the pen; corrected eye skips; and words or phrases that have been inadvertently repeated, then canceled.

Special collations are provided when a potentially authoritative text—for instance, a contemporary reprinting of uncertain origin—has not been used to establish the present text because it is probably derivative. The variants are recorded as a check on this decision.

R.H.H.

September 1977

University of California at Los Angeles