Contents Index

The Corrected Frankenstein: Twelve Preferred Readings in the Last Draft

David Ketterer

English Language Notes, 33:1 (Sept. 1995), 23-35

[Since this article was published, the notebooks cited therein have been published in facsimile as The Frankenstein Notebooks, ed. Charles E. Robinson, 2 vols. (New York and London: Garland Publishing, 1996) -- Ed.]

{23} A variety of errors, whether the responsibility of Mary and Percy Shelley, or of the typesetter, found their way into the original 1818 publication of Frankenstein by the London firm of Lackington, Hughes, Harding, Mavor, & Jones. Mary Shelley was certainly aware of textual defects; just eleven months after the January publication her Journal entry for 20 December 1818 includes the notation "Correct Frankenstein"1. Some of the errors survived into the revised editions of 1823 and 1831. In preparing his very useful edition of the 1818 text, James Rieger decided not to make any corrections: "The text printed by Lackington has been left intact, typographical errors and all. . . ."2 On the other hand, in the Mary Shelley Reader version of the 1818 edition, editors Betty T. Bennett and Charles E. Robinson explain that "we silently correct errors in spelling and punctuation. . . . All other emendations (e.g., the addition or deletion or changing of a word) are indicated by brackets in or footnotes to the text."3 Marilyn Butler takes the correcting process one stage further in her 1993 edition: "I have adhered to the spelling and punctuation of the original, but silently corrected misprints, and in two places restored missing words."4

In their 1994 edition of the 1818 text, D. L. Macdonald and Kathleen Scherf make yet more improvements in opting to "silently correct any typographical errors" and to correct six "misprints in the 1818 text that [Mary] Shelley corrected in the 1831 edition."5 These "misprints" include three likely typographical errors and three likely mistakes for which Mary Shelley and Percy (in so far as he read through, and corrected and revised, Mary's Last Draft and Fair Copy) were responsible. Macdonald and Scherf's implication that all of these "misprints" were first corrected in the 1831 third edition is mistaken. While two of them were first corrected in the 1831 edition, the remaining four had previously been corrected in a copy of the 1818 edition that Mary annotated, or in the 1823 second edition (which appears to have been the copytext for the 1831 edition).6

There is one final step to be taken before the 1818 text of Frankenstein will have been improved as much as humanly pos- {24} sible now that 144 years have passed since Mary Shelley's death. That step depends upon collating complete transcriptions of the Frankenstein manuscript materials in the Bodleian Library (the surviving Last Draft, which comprises almost all of the published text except for Walton's opening letters to his sister and the first nine paragraphs of Chapter 1 of Volume One, and the Fair Copy fragments, which comprise most of the three concluding chapters) and the 1818 edition. Since I have in fact completed those transcription and collation tasks, I am in a position (1) to indicate which substantive errors in the three editions of Frankenstein are traceable to the Last Draft and, if relevant, the Fair Copy, and (2) to propose seven substantive corrections sanctioned by the Last Draft additional to the four -- also, it turns out, so sanctioned -- which have already been made, making eleven preferred readings in all which should appear in an ideal critical text of the 1818 Frankenstein (and most of which should also appear in ideal texts of the two subsequent editions), and a twelfth which applies only to the 1831 edition.

For ease of reference, I relate page and line numbers (usually within parentheses) in Rieger's edition of the 1818 text to volume and page numbers of the Last Draft and to page numbers of the Fair Copy. The Last Draft, it should be noted, has frequently been misrepresented as the "rough draft"; in fact it approaches something like a rough copy (as opposed to a fair copy) of rough draft material which is now lost. It is organized as two volumes, separately paginated and chapterized. Only at the Fair Copy stage did Mary organize her text into the three volume form of the 1818 edition. It should be noted that the last thirteen of the 62 pages of the Fair Copy fragments are in Percy's hand.7

Four readily apparent lapses (three of which Macdonald and Scherf overlook in their corrected text) are traceable to the Last Draft. A first lapse traceable to the Last Draft occurs in the following sentence: Safie "sickened at the prospect of again returning to Asia, and the[n] being immured within the walls of a haram . . ." (Rieger 119.23).8 Macdonald and Scherf unnecessarily change "haram" to "harem" ("haram," as the OED indicates, is a legitimate variant)9 but they overlook the real error, the missing "n," which is also missing on the corresponding page 60 of Vol- {25} ume II of the Last Draft and must also have been absent in the no-longer-extant corresponding portion of the Fair Copy. In 1823 "and the being immured" was altered to "and being immured," and the alteration carried through to the 1831 edition.

A second lapse traceable to the Last Draft (and a second overlooked by Macdonald and Scherf) appears in this statement in the 1818 edition: "My wife and my sister will never recover their horror" (Rieger 134.16-17). The missing "from" here is also missing in the corresponding passage on page 80 of Volume II of the Last Draft and was not in fact inserted until the 1831 edition. All four of the collations of the 1818 and 1831 editions published to date fail to record this variant.10

The third mistake -- "Gower" (Rieger 157.11) -- is traceable to page 111 of Volume II of the Last Draft. William Godwin, Mary's father, or Mary herself corrected it to "Goring" (for George, Baron Goring, 1608-57) in the 1823 edition, a correction carried through to that of 1831.

A fourth readily apparent mistake (and a third overlooked by Macdonald and Scherf) is the first single form "was" in the following sentence: "Vegetables and bread, when they indulged in such luxuries, and even fresh water, was to be procured from the main land, which was about five miles distant" (Rieger 161.15-17). The same grammatical mistake appears on page 116 of Volume II of the Last Draft and must also have appeared in the Fair Copy. The plural form "were" should be substituted in an ideal critical text of Frankenstein, whether of the 1818, the 1823, or the 1831 editions. The same grammatical error appears in all of them.

There is a problematical just possible fifth mistake-traceable-to-the-Last-Draft to be considered in a quotation from Coleridge's "Ancient Mariner." Frankenstein describes how he "hurried on with irregular steps, not daring to look about me" [1.4.4]:

Like one who, on a lonely road,
Doth walk in fear and dread,
And, having once turn'd round, walks on,
And turns no more his head;
Because he knows a frightful fiend
Doth close behind him tread.

(Rieger 54.22-29)
{26} Mary Shelley seems here (and in the 1823 and 1831 editions) to be quoting from the original version of the poem published in the 1798 Lyrical Ballads. However, the opening line of the stanza quoted there reads as follows: "Like one, that on a lonely road . . ."11 Did Mary intentionally or accidentally change "that" to "who"? Interestingly, the Last Draft version of the stanza on page 79 of Volume 1, does not follow the 1798 Lyrical Ballads text; instead, it essentially conforms to that stanza in Coleridge's revised versions, published in the 1800, 1802, and 1805 editions of Lyrical Ballads; and to that stanza in the final revision published in 1817 in Sibylline Leaves, and in all subsequent editions of his poetry:
"Like one who on a lon[e]some road
"Doth walk in Scar and dread
"And having once turned round walks on
"Wal And turns no more his head
"Because he knows a frightful fiend
"Doth close behind him tread."
But the same "mistake" is present: "who" appears instead of "that". The first line of this stanza in the 1800, 1802, 1805, or 1817 texts reads as follows: "Like one, that on a lonesome road . . ."12

The question arises, should Mary's quotation from "The Ancient Mariner" be corrected to conform with the 1798 Lyrical Ballads "Ancient Mariner" stanza?13 I would answer "no." Frankenstein's apparent misquotation can be justified as indicative of the frailty of his human memory or as contextual paraphrase, but the possibility should also be considered that no misquotation is actually involved. Jim Mays, who has found approximately 300 new Coleridge poems in the twenty years of research that will eventuate in his three-volume edition of Coleridge's poems, has discovered, "Where there were 10 Ancient Mariners, there are now more than 100, a poetry editor's nightmare."14 Mary Shelley might well have had access to one or more of the ninety-plus other versions.

I turn now from mistakes traceable to the Last Draft to mistakes that may be corrected by the Last Draft. Of the twelve corrections sanctioned by the Last Draft, four were made, presumably by Mary, for the second and/or third editions of Frankenstein. {27} These mistakes may variously have been introduced at the Fair Copy stage (or at the proofing stage by Percy?) or they may be typesetter's errors. None of the relevant portions of the Fair Copy have survived. (1) The misspelling "Ingoldstadt" at Rieger 61.19 and repeated at 65.9, which Mary corrected to "Ingolstadt" in 1831, does not occur on the corresponding pages 92 and 100 of Volume I of the Last Draft. (2) The spelling "De Lacy" at Rieger 128.17, initiating a series replacing the previous "De Lacey." On the corresponding page 70 of Volume II of the Last Draft, Mary uses the form "de Lacey," and, in the pages following, the form "De Lacey" consistently appears. "De Lacy" was corrected to "De Lacey" in the 1823 edition and the corrections were carried into the 1831 edition. (3) The blatant omission "of my" from "the scene labours" (Rieger 161.9), does not occur on the corresponding passage on page 116 of Volume II of the Last Draft ("the scene of my labour"). Oddly, the same omission appears in the 1823 edition; the correction was not made until 1831. (4) The date "September 19th" (Rieger 213.27) should be "September 9th." ("September 9th" appears in the Last Draft on page 193 of Volume II.) The correct date appears in both the 1823 and 1831 editions.

There are eight preferred readings in the Last Draft which have not yet appeared in any published text of Frankenstein. Two of them have previously been proposed by E. B. Murray in his pioneering 1978 account of the Frankenstein manuscripts in terms of Percy Shelley's contribution.15 The first of these, in narrative order, is to be found in this passage on page 198 of Volume II of the Last Draft: "And do you think, said the dæmon that I was then dead to anguish & remorse?" On page 179 of Percy's corresponding portion of the Fair Copy, which is substantively what appears in the 1818 and all subsequent editions of Frankenstein, this becomes, "And do you dream -- said the Dæmon, do you think that I was then dead to agony and remorse" (cf. Rieger 217.19-20). Murray comments: "The italicized 'I' in the rough draft is rhetorically apt and probably should be restored as Mary's first and best intention."16 I agree.

Murray's second mistake-correctable-from-the-Last Draft appears in the fifth paragraph before the novel ends; the monster recalls that "I was nourished with high thoughts of honour and {28} devotion" (Rieger 219.65). The question begged, of course, is "devotion" to what? On page 200 of Volume II of the Last Draft, in place of "devotion" the compound "self devotion" appears; the corresponding page 183 of the Fair Copy, which is in Percy Shelley's hand, has "lf devotion." The fragment "lf" (i.e., [se]lf) appears because, in separating the leaves of notebook bifolia, the scissored edges of Percy's pages 181 to 186 cut off the ends (on the recto pages) or the beginnings (on the verso pages) of some of the transcribed words; only two pages (185 and 186) of a series of pages in Mary's hand, which presumably recopied these damaged pages for the printer, survive (Mary's duplicate copy starts in the paragraph immediately following the one here relevant and corresponds to Rieger 219.18-220.8).

Murray argues, I believe correctly, that "the progressive loss of self" culminating in the word's disappearance in the 1818 and all subsequent editions, came about because Mary, in recopying Percy's imperfect Fair Copy missed "one substantive . . . the curtailed 'self'."17 In claiming "self-devotion" here, the monster may be understood as reinforcing his doppelgänger relation with Frankenstein, whom he describes, a couple of pages earlier, as a "generous and self-devoted being!" (Rieger 217.23).

As a result of my transcription and collation work, five remaining errors in all of the three editions of Frankenstein, and a sixth in the third, can now be identified for the first time, and corrected from the Last Draft. I shall discuss them in order of ascending interest and importance.

There are two relatively minor emendations which should probably be made. On page 149 of Volume II of the Last Draft Frankenstein affirms "they all died by my hand[.]" In the 1818 edition "hand" appears as the unidiomatic "hands" (Rieger 182.22), perhaps as the result of a confusion with the alternative idiomatic form "at my hands." The same likely error appears in the 1823 and 1831 editions. As for the second minor emendation, it should be made in the unidiomatic verifying comment that Walton makes in his concluding communication to his sister: "Such a monster has then really existence" (Rieger 207.17). The corresponding line on page 184 of Volume II of the Last Draft includes two revisions: "Such a monster did ^has^ then {29} really existence"[.] The Fair Copy of this portion of the Last Draft has survived and on page 154 Mary apparently miscopied "really" as "really" and hence the error in the 1818, 1823, and 1831 editions.

There is a further minor emendation which I considered but finally rejected. On page 150 of Volume II of the Last Draft Frankenstein refers to "my journey to the Sea of Ice" (Mary's translation of the French place name "Mer de Glace" which is nowhere used in the novel). In the 1818 edition "Sea of Ice" appears as "sea of ice" (Rieger 183.33). The same lower case form appears earlier at Rieger 93.13 and on page 152 of Volume I of the Rough Draft. Both instances remain lower case in the 1823 and 1831 editions. Given the inconsistent manuscript evidence, the matter is arguable; since the place name does exist, there is a case for the capitalized form of the English translation in both instances. It seems more likely, however, that although Mary is referring to the Mer de Glace, she finally wished the phrase "sea of ice" to be understood as simply descriptive at both Rieger 93.13 and 183.33.

The first of the four more significant unidentified mistakes occurs in the following sentence related to Frankenstein's dilemma:

I wandered like an evil spirit, for I had committed deeds of mischief beyond description horrible, and more, much more, (I persuaded myself) was yet behind." (Rieger 85.7-9)
Exactly the same sentence appears in the 1823 and 1831 editions. In a note keyed to the word "behind," Leonard Wolf makes the following comment in his Annotated Frankenstein:
This is a puzzling locution. Its intent seems to be to say "much more was yet to come." Perhaps our author intended a metaphor as in "More, much more . . . was yet behind the veil of the future." (Wolf's ellipsis)18
Wolf could have spared himself this tortured hypothesis had he simply looked at his facing page reproduction of the corresponding Last Draft page (page 138 of Volume I). In place of "behind," the Last Draft has the words "in store." Wolf's oversight here is little short of extraordinary. Quite possibly Mary intended to substitute the word "ahead" as a synonym for "in store" and {30} mistakenly substituted the direct antonym of "ahead." Although Mary herself may have introduced the error in her Fair Copy (as a result of some mental aberration?), and although it was not corrected in either the 1823 or the 1831 editions, there can surely be no doubt that "in store" is the preferred reading.

Another rather peculiar unidentified error occurs in the following sentence related to the monster:

As night advanced, I placed a variety of combustibles around the cottage; and, after having destroyed every vestige of cultivation in the garden, I waited with forced impatience until the moon had sunk to commence my operations. (Rieger 134.35-135.4)
Again, exactly the same sentence appears in the 1823 and 1831 editions. In place of the nonsensical "forced impatience," the Last Draft, on page 81 of Volume II, has "forced patience." Once more, what Mary wrote in her Last Draft is clearly correct and what appears in all printed texts of Frankenstein is clearly a nonsensical error. Mary's Last Draft variant is again to be preferred and should be substituted in any ideal critical text of Frankenstein.

Before concluding with the perhaps most interesting of the more significant errors that can be corrected by the Last Draft -- two dating slips -- it is appropriate to note that another chronological pointer is not the error that one scholar has supposed -- in spite of its apparent correction in the Last Draft! On his return to Geneva after a prolonged absence at university, Frankenstein observes: "Six years had elapsed, passed as a dream . . . and I stood in the same place where I had last embraced my father before my departure for Ingolstadt" (Rieger 73.10-11). Leonard Wolf, in an annotation to this statement in The Essential Frankenstein (which is not in his original version, The Annotated Frankenstein), corrects the "Six years" to "five."19 There are two other similarly altered references to the same period of time in both the Last Draft and in all the published editions: "nearly five years" (Vol. I, p. 108) and "nearly six years" (Rieger 69.21); "Five years" (Vol. I, p. 120) and "Six years" (Rieger 75.27).

In the annotation immediately preceding his correction of "Six years" to "five," Wolf includes a chronology which purports to demonstrate that, in spite of Frankenstein's references to three {31} successive periods of two, or nearly two years (something of a creative tic on Mary Shelley's part since a fourth two year period will later be specified [Rieger 46.9, 52.25, 72.11, and later 151.15-6]), the actual elapsed time between Frankenstein's departure for Ingolstadt and his return to Geneva is closer to five years. Here then, "Mary Shelley creates a perplexity."20 She in fact does not; the elapsed time appears to be about five years and six months (the third two year period, a "nearly" one [Rieger 72.11], runs from the November of the monster's animation to the year after-next May of the servant Justine's arraignment for murder). In the Last Draft Mary rounded the figure down; for the 1818 and all subsequent editions, that figure was simply rounded up.

A real chronological error, the dating problem with regard to Frankenstein and Clerval's arrival in England, is apparent to any careful reader. At the end of Chapter 1 of Volume Two [Three], Frankenstein observes, "It was on a clear morning, in the latter days of December, that I first saw the white cliffs of Britain" (Rieger 154.31-32). In the next chapter he states that "We had arrived in England at the beginning of October . . ." (Rieger 156.24). Leonard Wolf notes that

Readers attempting to keep track of Victor's calendar must decide here which date to accept. For the sake of symmetry, this editor will continue to work with the October date. The choice of October seems reasonable too because it fits the assertion that Victor and Clerval spent some or several months in London.21
As I demonstrate below, Wolf made the right choice but in his appendix, "A Chronology of Events in Frankenstein," he reverses himself and has Victor and Clerval arriving in England at the "end of December."22 Chronological incoherence is clearly catching.

There are several date and time variants in the Last Draft of Frankenstein where the passage of time is advanced in the 1818 and subsequent editions. In this case, page 107 of Volume II of the Last Draft has Frankenstein and Clerval see the white cliffs of Dover "in the latter ^days^ of September" ("days" is an above-the-line insert), and the later page 110, like the 1818 and all subsequent editions, specifies "the beginning of October. . ." Since "the latter ^days^ of September is contiguous with "the {32} beginning of October," the Last Draft "September" is much to be preferred over the published "December." Clearly, "December" in the 1818 edition (at Rieger 156.24), and in the 1823 and 1831 editions, is a mistake which originated at the Fair Copy, typesetting, or proofing stages. The end of September date is consistent with Frankenstein and Clerval's previous activity. They were traveling down the Rhine "at the time of the vintage" (Rieger 152.28) before crossing Holland by post coach and then setting sail for England. As Wolf notes, "the time of the vintage" is "late August, early September."23 In other words, the journey down the Rhine and across Holland took a realistic three to four weeks.

The "December" in place of "September" mistake in the 1818 edition was compounded in that of 1831. In the Last Draft and in the 1818 and 1823 editions, Frankenstein and Clerval leave Geneva for England "in the latter end of August" (Volume II, page 102A [there is a previous page 102]; Rieger 151.15), a date consistent with the Last Draft arrival in late September. In the corresponding passage in the 1831 edition, "September" replaces "August" (Rieger ["Collation of the Texts of 1818 and 1831"] 253) presumably to accord more plausibly with the erroneous late December date of arrival. In the 1831 edition then the departure date should be corrected to conform with that in the Last Draft and the 1818 and 1823 editions (a fourth preferred reading of particular interest), and the arrival date should be corrected (as in the 1818 and 1823 editions) to conform with that in the Last Draft.

To the best of my knowledge, there are no further substantive emendations to the three editions of Frankenstein (or just the third) that can be justified in terms of the Last Draft (and/or the Fair Copy). In due course, seven of the emendations that I have here suggested (in two cases seconding Murray) will no doubt be made in an as-near-perfect-as-possible critical text of the 1818 edition; the eighth, of course, could only be made in a critical text of the 1831 edition. Mary Shelley's literary "progeny" will then be no less "hideous"24 but some of its worrying minor blemishes will have disappeared.


1. Mary Shelley, The Journals of Mary Shelley, 1814-1844, eds. Paula R. Feldman and Diana Scott-Kilvert, 2 vols. (Oxford: Clarendon, 1997) 1: 245.

2. "Note on the Text," in Mary Wollstonecraft Shelley, Frankenstein or The Modern Prometheus (The 1818 Text), ed. James Rieger (Indianapolis: Bobbs, 1974) xlv. (Rpt., U of Chicago, 1982).

3. Preface to The Mary Shelley Reader, eds. Betty T. Bennett & Charles E. Robinson (New York: Oxford UP, 1990) ix.

4. "Note on the Text" in Mary Wollstonecraft Shelley, Frankenstein or The Modern Prometheus. The 1818 Text, ed. Marilyn Butler (London: William Pickering, 1993) lviii. (Rpt., World's Classic paperback, New York: Oxford UP, 1994.)

5. "A Note on the Text," in Mary Wollstonecraft Shelley, Frankenstein; or, The Modern Prometheus. The 1818 Version, eds. D. L. Macdonald & Kathleen Scherf (Peterborough: Broadview P, 1994) 40.

6. Unfortunately, from the evidence of their edition, Macdonald and Scherf must have been unaware of E. B. Murray's "Changes in the 1823 Edition of Frankenstein," Library, Sixth Series, 3 (1981): 320-27. Murray argues that Mary's father, William Godwin, was most probably responsible for the 114 substantive changes tabulated and that Mary used the 1823 edition "as the basis of her copytext for her third edition revisions" (323). I think it likely that Mary was at least cognizant of three of the changes in the 1823 edition -- the normalizations of "De Lacy" to "De Lacey," the substitution of "Goring" for "Gower" and the date change from "September l9th" to "September 9th." Those are surely the kind of mistakes that she would quickly have become aware of after the book's first publication. I have relied on Murray's list of 1823 substantive variants (320-23) for dating the relevant Frankenstein revisions.

7. See Dep.c.477/1 (Volume I of the Last Draft) and Dep.c.534/1 (Volume 11 of the Last Draft), and Dep.c.534/2 ("the Fair Copy fragments) in The Abinger Deposit: Papers of P. B. Shelley, W. Godwin, and Their Circle (Bodleian Library, Oxford). For a full description of this material, see my article "(De)Composing Frankenstein: The Import of Altered Character Names in the Last Draft," which is forthcoming in the 1996 Studies in Bibliography. I am grateful to Lord Abinger for his kind permission to publish quotations from the Frankenstein manuscripts. And I am grateful to the Social Sciences and Humanities Research Council of Canada for a 1992-95 Research Grant which made it possible for me to create the first diplomatic transcription of those manuscripts.

8. All quotations from Rieger's text (see note 2 above) have been checked against the facsimile of the 1818 edition available in The Annotated Frankenstein, ed. Leonard Wolf (New York: Clarkson N. Potter, 1977).

9. Macdonald and Scherf, eds., Frankenstein 151.

10. In addition to the collations (230-59; 200-28; 317-59) in Rieger's, Butler's, and Macdonald and Scherf's editions of the 1818 text (see notes 2, 4, and 5 above), there is a select collation of the 1818 and 1831 editions in Mary Shelley, Frankenstein or The Modern Prometheus, ed. Maurice Hindle second edition (London: Penguin, 1992) 217-26. What is required, of course, is a collation of the 1818 and 1823 texts and then a collation of the 1823 and 1831 texts (see note 6 above).

Since Macdonald and Scherf's collation is the most complete and they claim to "list all the substantive variants between the 1818 and 1831 versions of Frankenstein" (317; my emphasis), it should be noted that they also omit the following five substantive variants: at 69.13 of their 1818 text, "my most" (as on page 2 of an insert to Volume I of the Last Draft) became "my" in 1831; a missing "Then" at 134.11 (it appears on page 7 of Volume 11 of the Last Draft) was restored in 1823 and the correction (also made in the Macdonald and Scherf text) carried into the 1831 text; "Turk" at 153.11 (and page 62 of Volume 11 of the Last Draft) mistakenly became "Turks" in the 1831 text; "Gower" at 186.34 (and on page 111 of Volume 11 of the Last Draft) was corrected to "Goring" in 1823 and the correction (also made in the Macdonald and Scherf text) carried into the 1831 text; and "September 19th" at 240.3 was corrected to "September 9th " (as it appears on page 193 of Volume 11 of the Last Draft) in 1823 and the correction (also made in the Macdonald and Scherf text) carried into the 1831 text.

11. Lyrical Ballads, ed. Michael Mason (London: Longman, 1992) 400.

12. Lyrical Ballads 198, line 440 (the 1800, 1802, and 1805 texts); Samuel Taylor Coleridge (The Oxford Authors), ed. H. J. Jackson (Oxford: Oxford UP, 1985) 60, line 446 (the 1817 text).

13. The following three recent Penguin editions of the 1831 text have in fact "corrected" the 1798 stanza. The 1994 Popular Classics edition and the Film and T.V. Tie-in edition both replace it with that which Coleridge first published in 1800 (both page 57). Maurice Hindle, in his 1992 edition, apparently substitutes the 1800 stanza with a variant opening line: "Like one, on a lonesome road who. . ." (58).

14. Nick Brooks and Tim Rayment, "Found: 300 Coleridge poems," The Sunday Times 12 February 1995: Section I, p. 3. Jim Mays informs me that he "has not so far discovered a version [of "The Ancient Mariner"] which corresponds to Mary Shelley's but this is not to say such a version does not exist" (letter to Ketterer, 13 March 1995).

15. E. B. Murray, "Shelley's Contribution to Mary's Frankenstein," Keats Shelley Memorial Bulletin 29 (1978): 50-68.

16. Murray 63.

17. Murray 64, 67.

18. Wolf, ed., The Annotated Frankenstein 125 n2. The manuscript page reproduction is on page 124. Alternatively, see Wolf's revision of The Annotated Frankenstein as The Essential Frankenstein (New York: Plume, 1993) 124, 127 n3.

19. Wolf, ed., The Essential Frankenstein 111 n27.

20. Wolf, ed., The Essential Frankenstein 110 n25.

21.Wolf, ed., The Annotated Frankenstein 333 n6; Wolf, ed., The Essential Frankenstein 217 n8.

22. Wolf, ed., The Annotated Frankenstein 242; Wolf, ed., The Essential Frankenstein 336.

23. Wolf, ed., The Annotated Frankenstein 226 n14; Wolf, ed., The Essential Frankenstein 211 n22.

24. "Appendix A: Mary Shelley's Introduction to the Third Edition (1831)," in Rieger, ed., Frankenstein (see note 2 above) 229.