University of Virginia Library

(De)Composing Frankenstein: The Import of Altered Character Names in the Last Draft by David Ketterer

The extant portions of the Last Draft and the later Fair Copy of Mary Shelley's Frankenstein are among the collection of Shelley and Godwin materials that Lord Abinger, their inheritor, deposited in the Bodleian Library in 1974 and 1976.[1] Since then various partial accounts of the Frankenstein manuscripts have been published, the most detailed and even-handed being "Shelley's Contribution to Mary's Frankenstein" by E. B. Murray. However, as Murray himself states in an initial footnote, he does not deal with all of those manuscripts: "I do not use manuscript materials I have seen only in microfilm . . ." (50 n1). The Last Draft reveals that Frankenstein, at the time of the Last Draft's creation, was conceived as two separately paginated volumes (as distinct from the three-volume Fair Copy, of which only concluding parts of the separately paginated Volume Three survive, and the original edition derived from it, issued as three separate volumes by Lackington, Hughes, Harding, Mavor and Jones in 1818). Murray does not report on the first of those Last Draft volumes. Furthermore, Murray appears to mistake the Last Draft — much of which clearly involved rough copying of material previously composed in rough draft form — for the original rough draft.

Like Murray, virtually all commentators on the Frankenstein manuscripts (notably, James Rieger, in his edition of the 1818 text, and Anne K. Mellor) have focused on the most sensational issue the material gives rise to: how substantial, how improving, how damaging was Percy Shelley's contribution to Mary's work? The Last Draft includes numerous corrections and additions, and some suggestions in Percy's hand, while the final thirteen pages of the last Fair Copy fragment, which are entirely in Percy's hand, contain significant revisions of the corresponding Last Draft section that Murray

assumes (I believe correctly) originated with him. The manuscript evidence in fact indicates that Percy contributed some 4,095 words, around 6% of the wordage of the 1818 edition — rather more than E. B. Murray's 2,200 estimate (50, 50 n1). Commentators have become so fixated on ascertaining the nature of Percy's input that virtually everything else that the Frankenstein manuscripts have to tell us has been ignored. It is my intention, then, in turning to that "everything else" to submit to particular analysis two of the seven instances in the Last Draft of altered or superseded character names and, taking into account relevant information in Mary's journal and one letter, elucidate what these changes suggest about the novel's order of composition/revision and the interpretative implications of that order.

I

It is necessary at the outset to provide, for the first time, a comprehensive description of exactly what the Frankenstein manuscripts consist of. It is rather surprising, given the two decades that they have been in the Bodleian, that nobody has put that basic information into print. The Bodleian designated the 1974 deposit Dep. c. 477/1, and the 1976 deposit Dep. c. 534/1 and Dep. c. 534/2. Dep. c. 477/1 and Dep. c. 534/1 are respectively Mary's Volume I and Volume II of the Last Draft. However those conceptual two volumes do not coincide with the two notebooks (one which contained light blue Continental paper, the other larger off-white British) in which the surviving draft was mainly written. Both notebooks were apparently intact at the time of writing and the leaves subsequently detached; the covers of both notebooks are now lost. The pagination of conceptual Volume II commences in the midst of the notebook that contained Continental paper. After the 77 surviving Continental leaves (60 Volume I leaves followed by 17 Volume II leaves), Volume II continues on the surviving 75 larger British leaves. Except for two rogue leaves described below, the Fair Copy fragments (Dep. c. 534/2) are written on leaves of British paper (bound at the time of writing with covers that are now lost) from what seem to have been two identical notebooks of a third type. They were probably modest exercise-type notebooks.[2]

In describing the manuscripts below, for ease of later reference, I have distinguished eight "natural" parts and identified them as A, A1, B, Cr, and D for the Last Draft; and, in the case of the Fair Copy fragments, DM, DP, and DPM. In relating these parts, and all manuscript quotations, to the

1818 text, I refer to page and line numbers in James Rieger's extremely useful line-numbered 1974 edition of that text (e.g., Rieger 103.5). In quotations from the manuscripts, carets are used to indicate an above-the-line insert.

The Last Draft — Volume I

A: The 121 pages (61 leaves) Mary numbered 41 to 161 (accidentally skipping 52, numbering two pages 58, and concluding with a blank page 161) (Rieger 30.13 ["servants had any request to make"] — 97.16 ["he thus began his tale."]). Assuming that all but the last leaf of the first two quires are lost, these surviving 61 leaves originally constituted the last leaf of quire II through to leaf eight (now apparently the sixth leaf since leaves two and three are missing) of the originally ten-leaved quire VIII (only quires V and VII contained twelve leaves) of a bound notebook of Continental paper. As usual Mary writes on both sides of each leaf. The light blue leaves (18.6 X 27 cm) of parts A and B, and probably Cr, all derive from the same notebook; its leaves were separated at some point after their being written on. The leaves originated as a quarto sheet; consequently the two watermarks — "D | ADIVONNE" and a bell — are split across the reconstructed folds. On each page a left hand margin of varying widths has been created by a pencilled line. A number of Percy's inserts and suggestions appear in these margins (as they do in the ruled and unruled margins of parts B — D). Taking account of the deducible pages 1 — 40, this part is sectioned by Mary into 15 chapters (with two chapter 7's).[3]

Bruce C. Barker-Benfield's reconstruction of the notebook quires indicates that leaves two and three of quire VIII (four pages before leaf 153 [3 altered from 7]/154) are missing because they were cancelled at the time of writing (Barker-Benfield 4 — 5). Following nine cancelled lines at the top of page 153 [3 altered from 7], a revision of the four and a bit cancelled pages (corresponding to Rieger 93.21 ["it rose"] — 97.16) continues through to the top of page 160 (approximately seven manuscript pages which deal with Frankenstein's meeting with the monster prior to the monster's central narrative).

A1: Insert material that appears on six unnumbered (by Mary) off-white pages larger than the Part A pages. Inserted at points on pages 43 and 44 of Part A above, it corresponds to Rieger 32.15 ("Natural philosophy is the genius") — 35.2 ("their place in my mind.") and to Rieger 35.17 ("I eagerly inquired") — 36.19 ("this various literature."). (Some of the first insert and all of the second were deleted in the revised 1831 edition.) A Bodleian librarian, the late Margaret Crum, in foliating in pencil all but one of the Volume I manuscript leaves (the accidental exception is Mary's pages 91 — 92), numbered these insert leaves 1, 2, 3, and 3v and then foliated Mary's Part A leaves (minus blank page 161) as 4 — 62, thereby giving the accidental impression that the insert material was written before the beginning of Part A. Part A1 must in fact have been written at some point while Part A was being written, or at some later stage of the drafting/revision process.[4] Part A1 consists of a folded sheet, a bifolium (with a rectangular portion torn from the bottom

of the second leaf some time before Mary wrote on it), followed by a single leaf of the same folio paper. All but the torn second leaf measure 20.2 x 31.6 cm. The centered watermark "T W & B | BOTFIELD" is visible on the leaves foliated 1 and 3. A centered crowned oval Britannia watermark incorporating the date "1815" is visible on the torn attached leaf 2. All the watermarks are upside down relative to Mary's script. Most probably these leaves of British paper were detached from the same folio notebook before being written on. There are no margins; Mary writes flush to the left and right hand edges of the page, on both sides of the leaves. This insert material is headed "Chapt. 2". It seems likely, therefore (see note 3 above), that a Chapter 1 portion of the missing forty pre-Part A pages corresponded to, or stood in place of, the frame narrator Walton's four prefatory letters and that Chapter 2 here corresponds to Chapter I (Volume One) of the 1818 edition.

The Last Draft — Volume II

B: The pages beginning on a verso headed "Vol. II" and paginated by Mary 1 ("161" — a continuation of the Volume I, or Part A, pagination sequence — appears on the blank recto) — 21 (corresponding to Rieger 97.17 — 109.8); they comprise manuscript chapters 1 and 2, and part of Chapter 3.[5] Most pages have pencil-ruled margins. The eleven leaves involved are of the same light-blue type as Part A and were originally part of the same quarto Continental notebook. They once constituted leaves 8 — 10 of quire VIII and the first eight leaves of quire IX (the remaining four are lost).

Cr: After a significant gap, a disjointed sequence (containing part of a deleted first and then a completed second rewrite of the now lost original Part [C]) of the same light-blue pages (most probably from the same Continental paper notebook as parts A and B) continues in an order that can be construed from the textual continuities and from matching leaves now separated. First are two separated bifolia consisting of: a recto numbered 57 (a cancelled

marginless page, here distinguished from a second recto numbered 57 as 57A), its marginless unnumbered verso (with the order of continuation number "59" about one-fifth down); a recto numbered 60 (a marginless page with the right hand bottom corner torn off before use), its verso numbered 61 (marginless with the corresponding left hand bottom corner torn off); a recto numbered 62 (marginless), its blank verso page; and a blank recto (with its right hand lower corner torn off before use), its verso numbered 63 (with an unruled margin and the corresponding left hand lower corner torn off). Bruce Barker-Benfield accounts for the two blanks by surmising "that MWS had accidentally turned over two leaves of the bound notebook" (9). Since the text is continuous between pages 61 and 62, these four leaves must at some point have formed the two central bifolia of their quire.

Part Cr continues with a bifolium which almost certainly, at some previous point, was the central bifolium of the same quire (it must have been detached for rewrite purposes before the continuous text of pages 61 and 62 was written). It was intact at the time of writing since penstrokes carry across the matched join. In essence, the first passage on what was this central bifolium replaces and continues the cancelled material on page 57A above and at the top of its unpaginated verso, while a second passage on the same bifolium continues from page 63 above. The once central bifolium consists of: a recto numbered 57 (57B, with an unruled margin and headed, by Percy, "another Chapter"), its verso numbered 58 (with an unruled margin); and a recto numbered 59 (with an unruled margin and text about one-third down the page identified by the continuation number "64"), its unnumbered verso lacking a margin and including text about one-fifth down the page identified by continuation number "65". The "65" material concludes (presumably because the two pages accidentally left blank had not been noticed, and because there was no more Continental paper) on a fragment of off-white, probably British paper (approximately one quarter the height of the white leaves of Part A1).[6]

The gap between 21, the last page of Part B, and 57, the first page of Part Cr, suggests, if the draft was paginated continuously, that 36 pages numbered 22 — 57 are lost. But only slightly more than eight pages account for the interval between the Rieger 1818 edition equivalent of the end of Mary's page 21 — (the appropriate) "how was that possible" (Rieger 109.8) — and the 1818 equivalent of the beginning of her page 57 — "Some time elapsed" (Rieger 117.17, the beginning of Chapter VI of Volume Two). It is not, however, an inevitable deduction that a substantial amount of text in the draft was cut; it is possible (indeed likely) that the first page number of Cr was made to coincide with a lost page of pre-Part D material and does not continue the 1 — 21 sequence of Part B. Since Part B ends five draft pages into Mary's Chapter 3, the missing pages (of greater or lesser extent) and the following extant pages would seem to have accounted for the remainder of Mary's Chapter 3 and a "Chapter 4", which, after what can be assumed to have been, perhaps on different occasions, deletions and rewriting, presumably corresponded to the 1818 Volume Two chapters V and VI. As will appear, the establishment of this Chapter 4 is important. After the rewrite stage that survives as Cr, it was divided into two chapters (presumably 4 and 5) presumably in line with Percy's direction at the top of page 57B, "another Chapter". His placement of this direction for the beginning of what presumably became Mary's "Chapter 5" corresponds to the beginning of the 1818 Chapter VI. That undesignated "Chapter 5," the rewritten second half of a long "Chapter 4," is all that now survives of "Chapter 4." I apologise to the perhaps wearied reader for what might seem a very finicky account of this central Part Cr, but it clearly relates to a portion of the manuscript that presented major problems and involved considerable revision. Under scrutiny, Cr provides the best clues as to the course of Frankenstein's redrafting and to an area of creative trauma.

D: The remainder of the Last Draft consists of seventy-five off-white folio leaves of British paper measuring 19.9 — 20.2 x 31 cm; they are, then, larger than the light blue leaves of parts A, B, and Cr, and very slightly smaller than the off-white leaves of A1. These leaves derive from five successive sixteen-leaf quires (originally in a bound notebook) of which only the second and third are complete (leaves 1 and 2 are missing from the first, leaf 16 from the fourth, and leaves 12 — 16 from the fifth). Three leaves cannibalized from elsewhere in the same notebook (and a scrap of different paper) have been inserted in quire II after its eighth leaf. The watermarked date "1806" below the initials "JL" is visible (upside down relative to the writing) at the center of 39 leaves, and a Britannia in crowned triple oval is visible (also upside down) at the center of 36 other leaves.[7] There are unruled margins throughout. (The columns of mathematical calculations in the margin of the first page appear to be basically two pagination totals.) The pages are numbered a hard-to-make-out 62 (possibly changed to 66 — the "2" is blotted in a way that makes it look like a "6", or like it may have been altered to a "6" to follow on from the Part Cr continuation numbers "64" and "65"; at any rate, the next page is 63) to 203 (Rieger 122.28 — 221.12 to the end).

After two cancelled lines, the first page of Part D, page 62, opens at the equivalent of the start of the 1818 Chapter VII (Volume Two). There is no chapter designation on page 62. Chapter headings appear for chapters 7 — 18 which are paginated 70 — 203.[8] Presumably the opening pages 62 — 69 constitute

the concluding portion of "Chapter 6." The opening portion of that "Chapter 6" is now lost. The also-now-lost Part [C] (as distinguished from Cr) seems to have constituted the end of a "Chapter 4." In other words, there is reason to suspect that the pagination and the chapter numbering of Part D was not the result of a continuous sequence initiated by Part B.

Barker-Benfield's reconstruction of the notebook quires indicates, as noted above, that three leaves were inserted after leaf 101A/102A (there is a second leaf numbered 101 and 102) and that one leaf (presumably two cancelled pages) is missing after leaf 181/182 which concludes with cancelled text (the following page number 183 appears to have been altered from 185). This evidence points to fourth and fifth significant areas of revision (additional to A1, the last two thirds of Volume I Chapter 14, and Volume II "Chapter 4") commencing at Rieger 149.5 ("But it is this gloom") and at Rieger 206.9 ("You took me on board").

The Fair Copy Fragments

My concern in this article is only incidentally with the Fair Copy (see foot-note 23 below) but for the sake of completeness a description of the Fair Copy fragments (which transcribe the latter portion of Part D above) follows. Unlike the very sketchy picture that we have regarding exactly when Mary began writing the Last Draft and when she finished, and in what order, the dates between which she (and Percy) transcribed the Fair Copy are recorded in her journal as 18 April — 13 May 1817 (I:168 — 169).

Except for two rogue leaves, the eight Fair Copy fragments are written on twenty-nine small quarto leaves, the smallest of all the manuscript leaves (18.6 x 22.4 cm). These small leaves were created by tearing or cutting bifolia in half (after their being written on since, in Barker-Benfield's words, "the writing on many leaves is torn or cut through" [16]). The two watermarks are each split across the reconstructed folds. The top half of a posthorn in crowned shield (a common watermark) is visible on six of these leaves and the bottom half with the appended monogrammed letters "P & S" on twelve others (all are the remains of a single-quire notebook of 24 leaves). The remains of a second single-quire notebook of 24 leaves consist of four leaves on which the watermark "PHIPPS & SON" is visible (which explains the "P & S") and seven others on which the balance of the same watermark is visible — the date "1809." The last leaf in Percy's Part DP bears the watermarked date "1814" set sideways at the inside edge jaggedly hacked just beyond the original fold; it derives from a third notebook, slightly smaller than the other two.[9] A second rogue leaf (see Part DPM below) derives from

a fourth notebook. All the Fair Copy paper is British and cream-colored. The Fair Copy fragments — their page numbers, beginning with 99, centered at the top of each page (instead of in the right and left hand corners as in most of the Last Draft) — correspond to the following three parts depending on who was transcribing:

This leaf was torn from a quarto notebook which was very slightly taller and wider than the two single-quire notebooks (based on what can be assumed from their surviving leaves). The top half of a posthorn in crowned shield watermark is visible set sideways at the torn edge.

It is apparent from the description above that, for a general sense of the rather complicated organizational relationships between the Last Draft and the 1818 edition, divisions created by physical factors (types of paper and a significant gap created by missing leaves) must be correlated with conceptual divisions in those versions. As a visual aid, the following chart relates those features including the framing divisions established by the narrators (W[alton], F[rankenstein], and the M[onster]).

My parts A — D factor in both the physical paper divisions and the conceptual Last Draft volumes. Parts A, B, and Cr apply to the Continental paper (ignoring here the Part A1 insert on British paper); Part D corresponds to the British paper. Parts A and A1 apply to Volume 1 and parts B, Cr, and D to Volume II.

II

Before proceeding further, it is necessary to relate an analysis of the Frankenstein Last Draft to the scanty information provided by Mary's journal entries and one letter — and to the quite divergent reconstructions of the composition process based on those documentary sources — in order to determine (1) whether the process of "composition" (which should be variously understood in relation to the Last Draft as including transcription, new composition, re-drafting, and revision) corresponded to the order of the published story, and (2) approximately what portions of the story were written when. I begin with a review of the documentary evidence.

Unfortunately, Mary's journal for the period 14 May — 20 July 1816, in the latter half of which period Frankenstein was begun, is missing. The anonymous Preface to the 1818 edition (written, according to Mary's 1831 Introduction, by Percy from his wife's point of view) simply refers to the

now familiar ghost-story compact (proposed by Byron according to Mary's more expansive Introduction to the 1831 edition) as occurring during "the summer of 1816 in the environs of Geneva" (Rieger 7). From information provided by Thomas Moore, Byron's biographer (confirmed by related entries in John Polidori's diary), it is clear that the compact must have been initiated at Byron's residence, the Villa Diodati (Moore 2: 31). Since Byron did not move into the Villa Diodati until 10 June 1816, and since his contribution, an unfinished vampire story, is dated 17 June 1816, the compact must have been proposed between those two dates. 16 June would appear to be the most likely date. In her 1831 Introduction Mary records that for several days she was unable to think of a story until, the night following a discussion about "the nature of the principle of life" (Rieger 227), she experienced the waking dream that inspired her: "On the morrow I announced that I had thought of a story. I began that day with the words, It was on a dreary night of November, making only a transcript of the grim terrors of my waking dream" (Rieger 228).[12] Since Percy Shelley and Byron left for a week's tour of Lac Léman on 22 June, Mary must have begun her story on that day at the latest — more probably a day or two earlier — in order to make her announcement.[13] The words she began with open Chapter IV of Book One of the 1818 edition.

In the 1831 Introduction, Mary writes, "At first I thought but of a few pages — of a short tale; but Shelley urged me to develop the idea at greater length" (Rieger 228 — 229). It seems reasonable to suppose that Percy made this suggestion around 22 June before he left for the tour of Lac Léman, or (less likely) shortly after returning on 30 June. Subsequently, at least the first line and probably other portions of the first composition stage of Frankenstein (the "transcript") were incorporated into a preliminary draft or preliminary drafts and eventually into the surviving Last Draft. Although the Last Draft might seem to constitute a second composition stage, it actually constitutes a third (one which incorporates the first and second stages).

In the surviving journal, which begins on 21 July 1816 (when Percy and

Mary and her stepsister Claire Claremont set off for a week's tour of Chamonix, a location which figures importantly in the novel), the first reference to Frankenstein appears (during the tour) in the entry for 24 July: "I read nouvelle nouvelles [Nouveaux contes moraux et nouvelles historiques (1802), by Madame de Genlis] and write my story . . ." (I: 118). Subsequent references to her writing (most probably Frankenstein) in Switzerland appear in entries for 29 July; 1, 5, 7, 9, 12, 16 — 22 (on the 21st she adds, "Shelley & I talk about my story" [I: 130]), 24, and 25 August 1816. Further references to her writing when back in England occur in entries for 16 and 18 September; 7, 17 — 27 (the last entry specifying, "Write Ch. 2 ½" [of the Last Draft Volume I or Volume II?] [I: 142]), and 28 October; 2 — 4, 6, 9, 11, 13, 15 — 18, 20 ("Draw and write (137)" [Mary's parenthetical number; I: 146] — presumably "137" refers to page 137 of the Last Draft but is it page 137 of Volume I or of Volume II?), 21 — 30 November; 1 — 11, and 13 December 1816; and 3 — 8, and 10 January 1817. A synoptic entry following 23 February 1817 and written, at the very latest, on 9 April 1817, includes this information: "the following week we enter our house [March 18] — Write every day" (I: 166). Presumably this means she wrote every day after moving into Albion House in Marlow and before her father William Godwin's visit on 2 April. The more than two month period after 10 January and before 19 March 1817 where no mention is made of any writing strongly suggests that the Last Draft had reached at least a first stage of completion by 10 January 1817. Since the entry for 10 April 1817 begins "Correct F." (I: 166) — the first of several such — it may be concluded that, after various second and third thoughts, perhaps prompted by Percy and other readers of her manuscript, Mary had certainly finished the Last Draft of Frankenstein before that date; most probably it was completed by 1 April, the day before Godwin's visit.

Two of the entries that I have parenthetically speculated about above call for further comment. The 27 October 1816 entry regarding "Ch. 2 ½" is followed on 28, 29, and 30 October, and on 2 and 4 November by references to Mary's reading Sir Humphry Davy's "Chemistry" (I: 142 — 144) — whether his Elements of Chemical Philosophy (1812) or his Discourse, Introductory to A Course of Lectures on Chemistry (1802; see Laura E. Crouch's argument for the latter), or both. If, as seems likely, this reading was undertaken to help with her writing at the time, the "Ch. 2 ½" reference must be to Volume I of the Last Draft. Most immediately, "Ch. 2 ½" might be identified as either the insert material headed "Chapt. 2" (Part A1) that deals with the contest between ancient "science" (as represented by Cornelius Agrippa, Albertus Magnus, and Paracelsus) and the modern science of chemistry, or, as might seem more likely, all of an original Chapter 2 (i.e., "3," altered from "2" [p. 47]) plus half of the next chapter, a chapter, originally undesignated, which later became Last Draft Chapter 4. Like Part A1, Chapter 4 (the last half of Chapter II of Volume One in the 1818 edition) deals with the contest between alchemy and chemistry (as it arises from Frankenstein's contact with professors Krempe and Waldman) (pp. 55 — 59; Rieger 40.19 — 43.30). Chapter number "2" in the Last Draft is altered to "3" in what appears to be the same

revision ink — the revision ink in which other changes on the same pages are made — as the designation "Chap. 4" (p. 55).

In pursuit of further insight into "Ch. 2 ½" as transcribed in the 1987 edition of Mary's journals, I examined the journal itself (Bodleian Dep. d. 211 [2]) and the actual reference. What Mary actually wrote was "Write Ch. 2 [altered from 3]½" (noted under "Words Obscuring Recovered Matter," Journals II: 700). The alteration is visible to the naked eye and beyond all doubt when examined under a magnifying glass. In spite of the fact that the "2" of Chapter 2 was altered to a "3" (Vol. I, p. 47) and that in the journal the reverse amendment apparently took place, the fact that the same numbers are involved makes it a well-nigh certainty that the confused "Ch. 2 [altered from 3]½" refers to "Chapter 3 [altered from 2]" plus its originally non-chapter-designated continuation. What became "Chap. 4" in the Last Draft is the half chapter. "Ch. 3 [altered from 2]½" could, of course, reasonably be understood as referring only to the half chapter following Chapter 2, in which case it denotes only what became "Chap. 4."

The common content strongly suggests that the Part A1 insert material was written with "Chap. 4" in mind and possibly shortly after "Chap. 4." If a chapter and a half, or even just the half chapter, comprising early pages of the Continental paper, were/was written/transcribed in just one day, it is likely that the preceding one or two chapters was/were written/transcribed during the two or three days immediately preceding 27 October 1816.

If what appears to be a page number in the entry for 20 November 1816 does indeed apply to a page of the Last Draft, it could refer to Part A page 137 of Volume I (corresponding to Rieger 84.16 ["may this be"] — 84.28]) or Part D page 137 of Volume II (corresponding to Rieger 174.14 ["neck, and screamed"] — 175.3 ["characterize that"]). The first page 137 marks the end of Chapter 11, Volume I of the Last Draft (and the end of Volume One of the 1818 edition) — it deals with Frankenstein's fiancée, Elizabeth, reconciling herself to the servant Justine's unjust execution. The second page 137, which is part of the account of Frankenstein's imprisonment in Ireland for murder, occurs several pages into Chapter 14, Volume II of the Last Draft (Chapter IV, Volume Two of the 1818 edition). Combined with the evidence regarding the placement of "Ch. 3 [altered from 2]½," it would seem most likely that the climactic Volume I page 137 is being referred to (making that page and the "Ch. 3 [altered from 2]½" pages the only passages from Frankenstein that can be dated with precision). The sad death of Justine might well have reminded Mary of the sad suicide of her half-sister Fanny Imlay on the ninth day of the month before she made the journal entry including the number "137."[14]

It would seem logical to deduce, then, that between 27 October and 20 November 1816, and most probably for a couple of days before and rather more after these dates, Mary was working on Volume I of the Last Draft,

that back in England Mary was transcribing/revising/composing on the Continental paper (whether that paper was acquired in England or, as is much more likely, in Switzerland), and that she had not transcribed, etc., any part of Volume II of the Last Draft before those dates — specifically during the early period back in England (say 16 September — 24 October 1816) — into a British notebook acquired shortly after her arrival in Bath. With regard to the last, however, one should allow for the seemingly less likely hypothesis, that in late October and throughout November she turned to Volume I after having written a substantial portion of Last Draft Volume II.

As I have indicated, 5 December 1816 was one of the days that Mary records doing some writing. A letter of the same day from Mary in Bath to Percy in Marlow contains what, as will become increasingly clear in the context of my reconstruction, is undoubtedly the most illuminating specific documentary reference to the writing of Frankenstein. It begins:

One should note here the phrase "the 4 Chap." instead of simply "Chap. 4." The article may imply that Percy would have aware of the material that she was referring to, perhaps because she had drafted an earlier version of it and that version was not satisfactory. Mary's stressing the length of this new chapter is probably related to Percy's now faint and hard to read pencilled marginal comment on the preceding Chapter 2: "This chapter is too short" (Vol. II, p. 11). In quoting the Chapter 4 reference in a note to Mary's 5 December 1816 journal entry, the normally perspicacious editors Paula R. Feldman and Diana Scott-Kilvert make an extraordinary mistake. They provide this clarification: "In chapter iv Frankenstein succeeds in discovering the cause of generation and life" (I: 148 n5). Ray Hammond in The Modern Frankenstein makes the same mistake regarding the same chapter allusion: "The fourth chapter of Mary's novel is the crucial scene-setting for Victor Frankenstein's creation. It is in this chapter that he describes his long work of raiding graves, visiting charnel houses and abattoirs to collect the materials for the creature he is to make" (170). Hammond, and Feldman and Scott-Kilvert, are all referring to Chapter IV of the 1831 edition (which is one of the shorter chapters in the book). They should, of course, be referring to Chapter 4 in the Last Draft, whether in Volume I or II. Chapter 4 of Volume I (which I have identified as the half of "Ch. 2 [altered from 3]½") is very very short (it corresponds, as previously noted, to Rieger 40.18 — 43.30). Mary must be referring in her letter to Chapter 4 of Volume II of the Last Draft. As I have conjectured in section I above, in its prior form this Chapter 4 (the last rewritten half of which survives as Part Cr) would presumably have

corresponded to chapters V and VI (the second corresponding to the rewritten Cr) of Volume Two of the 1818 edition — the story of Felix and Safie, the heart of the monster's narrative (itself the book's central narrative, framed as it is by Frankenstein's narrative, and the outermost Walton context).

It is possible that the unsatisfactory version of "the 4 Chap." that Mary, it may be, implicitly asks Percy to recall in her 5 December letter was transcribed (or drafted) in the latter pages of what could be the lost sixty-one pages of the British notebook. My hypothesis here is that for the Last Draft Mary began with the notion of drafting Volume I in the Continental notebook, and Volume II in the British notebook, but because the opening 61 and a bit pages in the British notebook was rethought, she followed Volume I in the Continental notebook with a revised version of all, or some of, the opening chapters of Volume II. In that case, Volume II "Chapter 4" in the Continental notebook would have replaced a series of pages (constituting one or more chapters) that ended ten pages before the surviving portion of "Chapter 6" in the British notebook.

Although there is no evidence of work on Frankenstein between 11 January and 18 March 1817, every journal entry from 10 April through to 17 April 1817 begins with the notation "Correct F" (I: 166 — 168). Entries for 18 — 25, and 29 April, and 3 and 9 May, record the act of transcribing the Fair Copy, culminating on 13 May 1817 with the words "Finish transcribing" (I: 169) — the entry for 29 April adding "and correct F" (I: 168). The last journal entry relevant to the composition of Frankenstein is for Wednesday, 14 May 1817: "S. reads hist of Fr Rev. [Précis historique de la Révolution française (1792) by Jean-Paul Rabaut Sainte-Étienne and Charles-Jean Dominique de Lacratelle] and corrects F. write Preface — Finis" (I: 169).[16]

There have been two very different attempts to reconstruct the process of Frankenstein's composition based on the facts assembled above: M. K. Joseph's "The Composition of Frankenstein," Appendix A to his 1969 edition of the 1831 text, and Emily W. Sunstein's much more expansive account in her biography of Mary Shelley, a 1989 winner of the Modern Language Association Prize for Independent Scholars.

Joseph narrows the starting date for Frankenstein to between 10 (or possibly 13) June and 22 June 1816, and goes on to point out that, sometime thereafter, Mary's original short story conception

The journal indicates further writing in late March not February but what is particularly significant about this account is Joseph's assertion that Frankenstein was "virtually complete" by late December 1816.

According to Sunstein, Frankenstein was only half done by that time. Her much more detailed reconstruction is characterized by a reliance on dogmatic assertion rather than argument; guesses are thereby passed off as facts. For Sunstein the novel was begun on the "morning after the poets sailed off" around the lake (122), i.e., on 23 June 1816. As I have indicated, the five sources — Polidori's diary, the 1818 Preface, Mary's 1824 letter to John Cam Hobhouse, Thomas Moore's 1830 life of Byron, and Mary's 1831 Introduction — do not allow for such precision. After the Chamonix tour, the Shelley party returned "to Chapuis [the Maison Chapuis, where they were staying] on July 27 . . . next day Mary observed the second anniversary of her elopement and continued her story" (Sunstein 123 — 124). Actually, Mary does not record any writing in her journal until the day after that, 29 July. After 21 August — the day "Shelley & I talk about my story" (Journals I: 130) — "Thanks to his 'incitement' she decided to make it into a full-length novel" (Sunstein 124). For Sunstein, this short-story version was written in Switzerland and the novel version after the return to England in September. As I have argued, I think it much more likely that the expansion decision occurred around 22 June (or shortly after 30 June). There is no logical connection between Mary's journal entry for 21 August and the conclusion that Sunstein draws from it. Having settled in Bath since to September 1816, "On October 18 Mary launched full into Victor's [i.e., Victor Frankenstein's] early life, and in ten days noted completing 'Chap. 2 ½' with his mother's death and his departure for Ingolstadt University" (Sunstein 127). Since the half refers to "Chap. 4" of Rough Draft Volume I, Mary had actually completed a bit more, namely her account of Frankenstein's conversion to modern science at the instigation of professors Kempe and Waldman. (And hence Mary's recourse to the work of Sir Humphry Davy around this time.) According to Sunstein, Mary "finished Volume I on November 20 . . ." (127). Sunstein does not explain that her evidence here is the "[137]" journal entry, and the Volume I she is referring to here is not, as one would expect, Volume I of the Last Draft but the rather different Volume One of the 1818 edition.

Mary, claims Sunstein, then switched to the "longer sheets" (127) that constitute my Part D of the Last Draft, in spite of the fact that chapters 13 and 14 of Volume I and chapters 1 — 4 of Volume II of the Last Draft intervene between page 137 of Volume I and Part D. Presumably she is assuming the existence of at least thirty-one longer leaves — 61 pages — now lost. "On December 5 Shelley left to look for a house at Marlow" and "Mary was finishing Chapter IV [of the Last Draft] of Volume II [of the 1818 edition]" (128). Here Sunstein has more or less correctly understood the reference to the "very long" Chapter 4 of Frankenstein in Mary's letter to Percy; she realizes (unlike Feldman and Scott-Kilvert, and Hammond) that Mary was writing about Chapter 4 in Volume II of the surviving draft, not Volume I.

Sunstein explains in a note that "This very long chapter was revised. The present Chapter VI of Volume II [of the 1818 edition] was originally V" (431 n37). Sunstein is in the right vicinity but, as I have explained, the revised draft "Chapter 5" (seemingly Percy's "another Chapter") or 1818 Chapter VI involved, was most probably previously part of a very long Chapter 4. She goes on to assert, "The morning after Shelley's return, December 15, . . . Mary was about halfway through Frankenstein . . ." (128). Sunstein has assumed, because Chapter VI of Volume Two of the 1818 edition marks roughly the halfway point, that the manuscript material which Mary drafted between 3 January 1817 (when Sunstein asserts, "Mary Shelley recommenced Frankenstein at about the point where the monster is driven off by the horrified De Lacys [a spelling at Rieger 128.17 which replaces the previous 'De Lacey']" [129]) and 9 April 1817 (the day Sunstein claims that Mary reached "the end of the last volume" [131]), corresponds to the second half of the 1818 edition, i.e., Chapter VII of Volume Two through to the end of Volume Three.

So, on the one hand, Joseph (who consulted a microfilm of what he calls accurately enough the "rough copy" [xv]) claims that Frankenstein was about done by late December 1816 while, on the other, Sunstein (who examined the actual Last Draft but apparently believed it to be the original rough draft of the novel) claims that Mary was only halfway there. Who is correct? For the moment I note only that Mary's extant journal records writing on 63 days in 1816 beginning on 24 July. To that figure must be added an unknown number of days between 22 June (at the latest) and 20 July 1816 (a period for which Mary's journal is missing). If Mary wrote every day during that period, her total of potential Frankenstein writing days in 1816 amounts to ninety-two (forty-eight of which apply to the period back in England). This plausible number should be weighed against the ambiguously recorded possible twenty-one days of potential Frankenstein writing in 1817. That figure includes the period when she claims to have written "every day" — presumably 19 March — 1 April 1817. That period was followed by a visit from William Godwin (2 — 6 April), and whatever dates in early April applied to "A[lba] comes. C.[laire] has been with us a week before" (Alba was Claire's baby daughter by Byron; Journals I: 166). The figures here are obviously in Joseph's favour.[17] It remains to determine, on the basis of altered character names in the Last Draft, the likelihood as to whether or not that draft was written, as Sunstein for the most part assumes, in the narrative order of the 1818 edition (only the initial Walton frame letters were added out of that order according to Sunstein [126]). If Sunstein's hypothesis

is false, it would be possible to account for Mary's writing a midpoint chapter in December 1816 and essentially completing the Last Draft of the novel around the same time.

III

Of the seven altered or alternate character names in the Last Draft of Frankenstein, two are of particular interest: Carignan/Clerval/Clairval and Amina/Maimouna/Safie. The first of these is in fact the first encountered in the Last Draft and in both cases the order I have given of the alternative names depends upon their appearance in the Last Draft narrative sequence — a sequence that may or may not have corresponded to the order of its writing.

In the second paragraph of Part A of the Last Draft are two references (the first introductory) to Frankenstein's long-term friend "Henry Carignan," with "Carignan" deleted and "Clerval" substituted: "Henry [Carignan cancelled] ˄Clerval˄ was the son[e cancelled] of a merchant [as cancelled] of Geneva [and cancelled] an intimate friend of my father[s cancelled]" (p. 41; see Fig. 1; cf. Rieger 30.27 — 28); "In [this cancelled] the description of our domestic circle I include Henry [Carignan cancelled] ˄Clerval˄" (p. 42; cf. Rieger 31.25 — 26). But a couple of sentences later, at the end of the paragraph, the name "Clerval" appears on its own as the unambiguous first choice: "we were never completely happy when Clerval was absent" (p. 43; Rieger 32.4). It is the form "Clerval," of course, which appears throughout the published text. The question arises: Was "Carignan" Mary's first choice for the name of Frankenstein's friend, in which case the deletions and the inserts represent her subsequent preference for "Clerval," a preference apparently fixed on at the end of the paragraph containing the second deleted "Carignan"? Or had she fixed on "Clerval" at some previous point in the composition process and experimentally or absent-mindedly substituted "Carignan" in the initial instances in Part A of the Last Draft but quickly changed her mind and reverted to "Clerval"? To determine which of these possibilities is the more likely it is necessary to track all the references to "Clerval" and "Clairval" ("Carignan" does not reappear in the entire extant Last Draft).

In the Part A1 insert material, the name "Clerval" appears in a passage which was deleted by a zigzag line: "[one evening cancelled] that I spent in town at the house of Clerval's [father?]" (folio 3r, as numbered by Margaret Crum of the Bodleian; the entire omitted material follows on "cheerfully consented" at Rieger 36.2).

After the initial page 43 instance an unsubstituted "Clerval" appears twenty more times in Part A of the Last Draft: three times on page 80 (Rieger 55.2, 55.6, 55.12), once on pages 83, 84, 86, and 87 (Rieger 56.24, 56.35, 57.33, 58.9), twice on page 96 (Rieger 62.32, 63.11), once on page 97 (Rieger 63.22), three times on page 98 (Rieger 64.2, 64.10, 64.15), twice on page 100 (Rieger 65.2, 65.12), once on pages 101 and 106 (Rieger 65.20, 68.20), twice on page 107 — the first time in a Percy insert (Rieger 68.29, 68.32), and on page 108 (Rieger 69.12).

After Frankenstein bids farewell to Clerval following the last reference

above — in the penultimate chapter of Volume One of the 1818 edition — Clerval does not reappear in that edition until the opening chapter of Volume Three (at Rieger 150.18). But in the corresponding Last Draft Chapter 10 of Volume II (Part D) he first reappears as "Clairval"; Frankenstein's father suggests, in a passage cancelled by three vertical lines, that "you should accompany Henry Clairval in a journey he intends to make to England" (p. 100; cf. Rieger 151.5 — 6). This prompted the following marginal suggestion by Percy, a suggestion cancelled (because acted upon) by a vertical line: "I think the journey to England ought to be Victor's proposal: — I think he ought to go for the purpose of collecting knowledge for the formation of a female. He ought to lead his father to this in the conversation, — the conversation commences right enough" (p. 100). The major rewrite which this prompted, the fourth such in the extant Last Draft, appears on three inserted leaves following leaf 101/102 (or for citation purposes 101A/102A): 101B/102B, blank/unpaginated; unpaginated/unpaginated. All these inserted leaves are probably from the same notebook (the first two leaves are an intact bifolium). The rewrite corresponds to Rieger 149.5 ("But it is") — 151.14. Suddenly, the spelling "Clerval" appears five successive times: twice in a cancelled passage on page 101A, once in the inserted rewrite on the first unnumbered recto (Rieger 150.18), and twice on page 103 (Rieger 151.34, 152.8).

But all the following references in Last Draft Chapter 11 (albeit "11" is deleted and replaced by "2," the chapter number in the 1818 edition Volume Three) through to Chapter 17 — twenty-two in all — are to "Clairval" corrected as "Cl[air cancelled] ˄er˄val" or "Cl[ai cancelled] ˄e˄rval."

When Frankenstein's friend is next referred to — in Chapter 18 of Volume II of the Last Draft (the last Part D chapter) — it is after his death and he appears once more as "Clerval": in his sleep Frankenstein "beheld Clerval enjoying health & youth" (p. 177; cf. Rieger 202.7). And the remaining three references follow suit (p. 183, Rieger 206.26; p. 187, Rieger 209.18; p. 188, Rieger 209.32).

In the Last Draft the various references to Frankenstein's friend appear in the following sequential order of variants: [Carignan cancelled] (in Part A) and Clerval (in parts A and A1); and Clairval, Clerval, Cl[air cancelled] ˄e˄rval or Cl[ai cancelled] ˄e˄rval, and Clerval (all in Part D). If this order of variants does in fact correspond to the order of writing, one must attribute the bizarre order to extreme forgetfulness or inconsistency.

Whatever might be imputed to the order Carignan-Clerval-Clairval, there is, however, at least one readily recoverable logic to the sequence Clairval-Clerval-Carignan. It could represent an increasing movement away from Mary's likely source, the surname Clairmont as in Mary Jane Clairmont, her stepmother, Charles Clairmont, her stepbrother, and Clara Mary Jane Clairmont (or Claire Clairmont as she preferred to be known), her stepsister. Claire was Mary's co-dweller (along with Shelley) at the Maison Chapuis. Mary could well have arrived at the name "Clairval" by substituting the French "val" (valley) for the French "mont" (mountain), a feminine symbol for a

masculine one.[18] (Further support for this derivation is provided by the fact that Claire, Byron's insecure mistress, in a letter informing him that she was on her way to Geneva, told him to address letters to her as "Clairville" so as to distance herself from the name "Clairmont" and ugly memories of her mother [Grylls 59 n3].) It would seem logical to suppose that Mary subsequently altered "Clair" to "Cler" in order to make the connection between "Clairval" and "Clairmont" less blatant.

Considered in the light of the hypothesis that "Clairmont" was a move away from "Clerval," Mary's momentary recourse to the name "Carignan" may not have been her starting point; rather it may be evidence of a later impulse to obliterate entirely what might too easily be seen as an allusion to the name "Clairmont." A more likely hypothesis, however, would be that "Carignan" was the name that Mary originally used in the now lost rough draft, or rough drafts. Its brief appearance at the beginning of the Last Draft would represent a momentary confused harking back to that earlier version because the Last Draft was being copied from the earlier version. In either case, it seems most productive to focus attention on the Clairval/Clerval variants.

The most plausible model for Clerval is not Mary's stepsister but her stepbrother Charles Gaulis Clairmont. Born in 1795, in 1816 he would have been much the same age as Clerval when the plot of Frankenstein begins. Clerval displays Charles' talent for languages. Indeed "Languages were his [chief cancelled] ˄principal˄ study for he wished to open a field for ˄their˄ self-instruction on his return [to Geneva] — [Greek cancelled] Persian [& cancelled] Arabic & Hebrew gained his attention [when cancelled and followed by Percy's "X" keyed substitution X as soon as] he had become perfectly master of the Greek & Latin languages" (Vol. I, p. 99; cf. Rieger 64.16 — 20). In The Godwins and the Shelleys, William St Clair writes of Charles, "As befitted a child from such an international background [his mother had French family connections and his father Karl Gaulis, anglicized to Charles Clairmont, came from a noted Swiss family] he showed an early talent for languages, speaking French and German as well as later learning Spanish and Italian" (297). He would work as an English language teacher in Vienna and, in 1828, hoped (futilely) to be appointed professor of German at the University College of London.[19]

If, discounting "Carignan," the surviving Last Draft of Frankenstein was written according to the sequence "Clairval," then "Clerval" (and the "Clairval"s then corrected), Mary must have begun with Part D on the British paper (the 1818 Volume Two, Chapter VII — Volume Three, Chapter VII). In other words, in Part D she first decided on "Clairval" at the beginning of Chapter 10, toyed with "Clerval" in the remainder of Chapter 10 (apparently both before and after Percy's read-through and his re-write suggestion), returned to "Clairval," and finally fixed on "Clerval" in Chapter 18; then, in Part A she continued to use "Clerval," except for a brief, perhaps absentminded, initial substitution of "Carignan."[20] However, if Mary was to some

extent alternating between Volume I in the Continental notebook and Volume II in the British notebook, it is possible that Volume II Chapter 18, with its "Clerval" references, was written after the establishment of "Clerval" in Part A.

At least consistent evidence for this reconstruction of the order in which Mary Shelley wrote the surviving Last Draft — essentially a process of circling back — is provided by the Amina/Maimouna/Safie sequence of variants. These variants follow five other altered character names in the Last Draft which occur in the following order: Waldham/Waldman, Justine Martin/Justine Moritz, Miss Mansfeld/Miss Mansfield, Hofland/Duvillard, and the particularly teasing Elizabeth/Myrtella/Elizabeth. While all these are of interest in their own right, they will not help us with the task of de-composing Frankenstein and hence my discussion of them is relegated to an appendix.

The character finally named Safie figures in the monster's narrative, the central of three concentric or nested narratives (Walton's, Frankenstein's, and the monster's). In an improbably contrived situation, while eavesdropping on the inhabitants of a cottage (an old blind man named De Lacey and his grown children, Agatha and Felix, and Felix's fiancée Safie), the monster learns to understand and speak French (as Felix teaches the Turkish-born Safie the language), reads three books (Paradise Lost, one volume of Plutarch's Lives, and The Sorrows of Young Werther), and learns of Safie's history.[21]

After the missing pages 22 — 57 (which would have included the introduction of Safie), in the complicated, jumbled text of Part Cr that Percy headed

"another Chapter" (57B; Chapter VI of Volume Two of the 1818 edition), there are three references to "Safie" which also occur in the 1818 edition: two on page 57B (Rieger 118.3 and 118.8) and one on page 58 (Rieger 118.24), and a fourth which does not (in an abandoned sentence fragment): "The Turk informed Safie of his intentions and" (p. 59; following Rieger 118.33). A horizontal line has been drawn under this sentence fragment; and under the line, and against the continuation number "64" in the margin, the text continues from page 63. It further continues on the unpaginated verso of page 59, which has the continuation number "65" in the left margin about one-fifth down the page, and concludes on one side of an unpaginated fragment of white paper, probably British (for exact correlations with the 1818 edition see footnote 6 above). Five more "Safie" references appear in this textual hopscotch: one in the "64" material (Rieger 121.26), and four in the "65" material (Rieger 122.7, 122.20, at 122.21, and 122.24), the third being deleted in favour of the insert "˄the Arabian˄." All these "Safie references occur on the pages of the third or central Part Cr bifolium.

In a previous version, on the cancelled page 57A there is one reference to "[the Arabian cancelled]" that was changed to "Safie" in the rewrite on page 57B (the first such reference noted above); the immediately following "Safie" reference in the revision is missing in the concelled version. That version breaks off with "in his favour" (unpaginated verso of 57A; Rieger 118.21) and, after a horizontal line, a lengthy passage commences on the same physical page under the continuation number "59" in the left margin. This passage, which follows on "the consummation of his happiness" on the recto numbered 59 (Rieger 118.33), begins with the words "During the ensueing days" ("59"; Rieger 118.34) and ends with the words "deprived of his wealth" (p. 63; Rieger 121.24). It is in the first paragraph of this "59" passage that the Amina/Maimouna/Safie variants first appear. Since "Safie" was the third chronological choice, the conclusion is inescapable that the "59" passage, like the cancelled passage it follows on, was written before the passages containing the unequivocal "Safie" references detailed above.

In the "59" passage opening paragraph, the monster recalls that "[Amina cancelled] ˄Safie˄ related that her mother was a Christian Arab" who "had won the heart of the [pr cancelled] father of [Amina cancelled] ˄Safie˄" (cf. Rieger 119.13 — 15) and (continuing on page 60) that "This Lady died, but her lessons were indelibly impressed in the mind of [Maimouna cancelled] [˄Amina˄ cancelled] ˄Safie˄" (see Fig. 2; cf. Rieger 119.22). At the conclusion of this essay I shall explain Mary's choice of the names "Amina," "Maimouna," and "Safie"; but for the moment it is essential to note only that in this paragraph "Amina" was Mary's first chronological choice, "Maimouna" her second, and "Safie" her third.

The two appearances of "Amina" here are as puzzling as the two earlier appearances of "Carignan." Both deleted names disappear entirely from the surviving draft after their second appearance. And in both cases Mary could be accidentally copying superseded name choices from her earlier, now lost, rough draft(s). There may well have been a preliminary version of the novel

in which Frankenstein's friend was called Carignan and Felix's fiancée was called Amina.

In Part D of the Last Draft there are six instances where Mary's apparent second choice "Maimouna" appears; in each case "Maimouna" is deleted in favour of the inserted "˄Safie˄" (p. 67 [Rieger 126.35], p. 70 [Rieger 128.21], twice on p. 75 [Rieger 131.23 and 25], 82 [Rieger 135.26], and p. 184 [Rieger 207.13] in the closing Walton frame).[22] The conclusion is again inescapable. Since a cancelled "Maimouna" replaced by a "Safie" insert appears in the last Part D chapter, virtually all of Part D must have been written before at least all the portions of Part Cr where the name "Safie" appears not as an insert but as Mary's clear final name choice — something which happens at least once somewhere on every page (including the scrap) except for the three first written (pages 57A, its unpaginated verso with the continuation number "59," and page 60). In fact, since this rewritten material, however erratic, appears to have been drafted consecutively, we may conclude that all of Cr must have been written after virtually all of Part D; the Safie variants in the first paragraph of what I have called the "'59' passage" are all accidental throwbacks to, or copyings of, superseded name choices. This could well apply even if the change of ink and pen-cut at the start of the page 60 paragraph immediately following the Safie variants indicates an interval elapsed after Mary's decision to temporarily break off at that point and rewrite the first rewrite (see Fig. 2). What we have here, then, is a circling back movement, similar to the movement suggested by the Clairval/Clerval variants from Part D to parts A and A1.[23]

IV

Putting these two movements together implies a copying/redrafting process that might be crudely imaged as an inward-moving spiral. However, before arriving at any final conclusions it is necessary to consider all the content evidence in relation to all the physical and all the temporal evidence. The fact that Part Cr was drafted after all, or virtually all, of Part D does not prove that the previous material on Continental paper was written after Part D. There are, however, three additional matters of content which do further indicate that most, if not all, of Part D was written before all, or at least some very substantial portion, of the material on Continental paper.

The first of these involves the assumption that, if Part D was written first, it might well contain temporal pointing that would be disjunctive with that in the preceding material. The matter of Clerval's age is consistent with this assumption. Simply put, the chronology established in parts A through Cr puts Clerval's age at the time of Part Cr around 22, while in Part D, after the passage of two more years, his age is estimated at around 20 (instead of around 25).

When Clerval is introduced at the beginning of Last Draft Volume I, it is implied that he is the same age as Frankenstein: Clerval "compensated" for the fact that "My brothers were considerably younger than myself . . ." (p. 41; Rieger 30.25 — 26). Frankenstein is "seventeen" (p. 47; Rieger 37.9) when he goes to university; after working "hard for nearly two years" (p. 76; Rieger 52.25) he animates his creation; when he finally returns to Geneva "nearly five years" (p. 108) have passed ("nearly six years" in the 1818 edition [Rieger 69.21]) since he left, nearly "two years" (p. 113; cf. Rieger 72.11) since the monster's animation — "Five years ["Six years" (Rieger 73.10)] had elapsed — passed as a dream . . ." (p. 115), a time notation mentioned a third time on page 120 (cf. Rieger 75.27). (The three changed references to "six years" in the 1818 edition represents a rounding upwards rather than downwards; the elapsed time in both the Last Draft and the published work is actually about five and a half years.) About "two months" (p. 145; Rieger 89.20 — 21) after the wrongful execution of the servant Justine, Volume I ends with the monster about to tell his story. At this point both Frankenstein and Clerval must be at least 22.

But in Volume II Part D (which continues the monster's narrative begun in Volume II parts B and Cr), after a fourth two-year, or nearly two-year, passage of time, "two years of exile" (p. 102A; Rieger 151.15 — 16), Clerval's corpse is described by a witness as that of "a handsome young man about twenty years of age" (p. 134). If Mary was writing in the British notebook with an awareness of the previously written, careful and reiterated time specifications in the Continental notebook, she would surely have calculated that Clerval at his death would be about 25 years old. The mistake is understandable however if Mary had not yet worked out a detailed chronology for Volume I of her Last Draft because at least the extant portion of it had not yet been redrafted. In the 1818 edition, taking account of the six year absence, Clerval's estimated age at his death is corrected to "about five and twenty years" (Rieger 172.19).

A second area of content evidence conclusively demonstrates that Part Cr was written after at least the opening of Part D. The end of Part Cr overlaps (rather than underlaps) the beginning of Part D. Part D begins on page 62 as follows: "[having overcome many difficulties succeeded at length in joining her lover in his retreat cancelled] [¶] Such was the history . . ." (cf. Rieger 122.26 — 28). The sentence which is not cancelled, but which appears to have been written immediately after the cancelled fragment, corresponds exactly to the opening sentence of Chapter VII of Volume Two of the 1818 edition. A passage that constitutes a rewrite of the two cancelled lines occurs

in Part Cr on the unnumbered fragment of paper that concludes the material keyed to continuation numbers "64" and "65": ". . . after her death the woman of the house in which they had lived took care that Safie should arrive safely at the cottage of her lover." This corresponds to the closing sentence of the preceding Chapter VI of Volume Two of the 1818 edition (Rieger 122.25 — 27). It follows as a certainty that two opening lines of Part D and at least some of the contemporaneously written, immediately following material was drafted before Cr. Furthermore, it may be conjectured that Cr replaces a cancelled section of text (in the British notebook?) that began on page 57 line 1 and continued through to page 62 line 2.

Although it cannot be conclusively proven that most or all of the rest of Part D, the rest of the extant material on British paper, was redrafted before all of the extant material on Continental paper, there is a third and final point related to content (and to the second point above) that certainly further encourages that conclusion. I have raised the possibility that the pagination of Part D does not continue the Volume II pagination begun by Part B. Correspondingly, the chapterization of Part D appears not to follow on that begun by Part B and continued by the missing long "Chapter 4." The following diagram attempts to crudely relate the adjoining chapterizations leading up to and away from the area of overlap between text on Continental paper (including the concluding slip of different paper) and that on British with the chapterization of Volume Two of the 1818 edition. The first line represents material assumed to have been written earliest in time; the last (the 1818 text) that created latest. The broken lines represent material that can be assumed but is not extant.

It is possible that the material on British paper is here out of the correct chronological order and should appear in between the very long "Chapter 4" and "another Chapter." If that were the case, "Chapter 4" must have been converted into chapters 4, 5, and 6 (which may have been written on British paper? or corresponded to the same numbered chapters on the British paper?), rather than into the two chapters that the surviving evidence would lead one to suppose. Only then would the chapterization on the British paper, which commences with "Chap. 7th" (p. 70; written at the same time as the chapter opening) follow on that which preceded in narrative sequence.

It might be common-sensically supposed from the physical evidence that Mary wrote on the Continental paper while in Switzerland and subsequently on the British paper when she was back in England. However, the evidence that Mary Shelley's allusions in her journal to "Ch. 2 [altered from 3]½" (on

27 October 1816), and to "(137)" (on 20 November 1816), and in a letter to "the 4 Chap." (on 5 December 1816) are all to be applied to sequential portions of parts A, [B — C], and [C] (where [B — C] and [C] represents the sections of the Last Draft missing between parts B and D) comes close to proving that, when she was back in England, she was writing on the Continental paper. The spaced narrative order of the allusions strongly suggests that, from approximately the last week of October to the first week of December 1816, she was writing parts A, A1, B, [B — C], and [C] of the Last Draft. The question remains, If Mary was writing parts A through [C] after being back in England for six weeks or so, what Frankenstein material was written in Switzerland and the first weeks back in England? Although three allusions, apparently to content aspects of the Last Draft on Continental paper, have survived, unfortunately there are no surviving allusions that can be linked to Mary's writing in the British notebook.

One might speculate as to whether or not the five and a "half" chapters which at some point presumably preceded Part D were Last Draft chapters that corresponded to the opening chapters of the rough draft of Volume II, or were in fact an entire preceding version of Frankenstein — a version that had not been divided into two volumes. Such a version might have begun with the animation of the monster and only taken five and a "half" chapters (amounting to 61 pages) to reach the point at which Part D begins. Five and a "half" chapters does seem about right for 61 pages; 62 more pages brings the narrative to the end of Chapter 12 on page 123 (i.e., six and a "half" chapters in only slightly more pages). To understand Part D as the Last Draft continuation of an opening portion that would be discarded, could account for the apparent chapterization discrepancy between parts A to Cr and Part D. It could also account for the discrepancy between the apparent thirty-six-page manuscript gap between parts B and Cr and the only just over eight seemingly corresponding pages in Rieger's 1818 edition. If there was a one-volume version of Frankenstein, of which only Part D now survives, that version would have contained eighteen chapters, one for each year of its young author's age at the time the work was conceived and its composition begun.

But to return to somewhat harder evidence, what of the temporal data? There is evidence of approximately fifty Frankenstein writing days before the last week of October 1816, sufficient time to produce a preliminary draft of the novel in Switzerland and to produce on British paper the Part D "rough copy" during the first six weeks back in England (which include approximately twelve recorded writing days). It seems rather unlikely that Mary drafted Part D in the first quarter of 1817 prior to 9 April, the latest day on which she might have completed the Last Draft. There is little evidence of sustained composition during this period. The data is much more compatible with revision work.

In the light of all the content, physical, and temporal evidence there would appear to be at least two possible timetable reconstructions of the process of rough copying/redrafting based on the following two possible

approximate orderings of the Last Draft parts [including lost parts within square brackets]: A and A1, B, [B — C], [C], D, [B — Cr?] and Cr (where [B — Cr?] represents a revision of [B — C], if there was such a revision, and Cr of [C]); or [pre-D?], D, A and A1, B, [B — C], [C], [B — Cr?], and [Cr] (where all, or some portion of B — [C] represents a revision of [pre-D?], and Cr survives as a second revision of the last part of [pre-D?], the lost 61 pages [if it was 61 and a gap had not been deliberately left] that preceded the present Part D). Given, under content evidence, (1) the lack of references to "Clairval" (in place of Clerval) in parts A and A1, and the indecision between "Clairval" and "Clerval" in Part D, (2) Clerval's regressed age in Part D, and (3) the possible chapterization discrepancy between the two notebooks, the second ordering, involving a circling back to parts A and A1 followed by movement inwards to Part [Cr], seems the most likely. The temporal evidence pointing to a lack of sustained Frankenstein composition in 1817 supports this order. A physical evidence argument also in favor of this second ordering would note the oddity of there remaining at least three unused bifolia of Continental paper on which Cr would be written after the completion of Part D, if one were to assume that the drafting of Part D directly followed on the missing Part [C], i.e., that Mary moved on to the English notebook when she had completed the Continental one or otherwise run out of Continental paper.

The following four stages of composition, redrafting, and revision would be compatible with my second ordering:

• (1) The lost "transcript" of Mary's "waking dream" beginning: "It was on a dreary night of November". This was probably written somewhere around 22 June 1816. It may have led almost immediately to her jotting down an outline for a "short tale" (Rieger 228).
• (2) A novel-length version in preliminary rough draft or rough drafts form. This may have been written in medias res as a continuation of the Stage 1 "transcript" before backing up to include earlier material. It may have been composed in the form of separately paginated segments.[24] This preliminary

  ---

  263

  
  
  
  version (and perhaps the first chapter of Volume I of the Last Draft) could have been written in Switzerland between somewhere around 23 June and somewhere around 25 August 1816.
• (3) The Last Draft. It is possible that Volume I on Continental paper and Volume II on British paper were, to some degree, written concurrently. Possibly having written Chapter 1 of Volume I (whether in Switzerland or in England) Mary Shelley in Bath, knowing that the following chapters presented difficulties and would require some scientific and philosophical reading (for her account of the monster's growing awareness she studied Locke's An Essay Concerning Human Understanding intensively between 18 November 1816 and 8 January 1817[25]), may well have turned to redrafting Volume II, which begins with the impetus of the monster's narrative, in a perhaps newly acquired British notebook. The copying/redrafting of Volume II, or perhaps of all but the very long last chapter of Volume II if she had

  ---

  264

  
  
  
  not yet fully figured out the frame situation, could have occupied her between approximately 16 September and 24 October 1816.[26] Mary then, around 25 October, could have circled back to Volume I, picking up the narrative at the point she might have abandoned it (after Chapter 1?). She then completed Volume I and moved on, using the same Continental paper, to a revised drafting of the opening pages of Volume II (hence the first 61 pages drafted on British paper were discarded) completing Part [C] by 5 December 1816.
• (4) The revised Last Draft. The opening frame letters and all or part of the closing frame chapter and various local revisions, including the major revision Cr (and a possible [B — Cr?]), could have been made during the twenty-or-so writing days of the period January to approximately 1 April 1817.

Alternatively, my first ordering of the Last Draft parts — to my mind less likely — would involve some rejigging of stages 2 — 4 above. The preliminary drafting would need to have continued back in England until the last week or so of October. The material on Continental paper would then have been copied/redrafted during the period late October to 5 December 1816, after which, through to perhaps 10 January 1817 or some date beyond, Mary continued and concluded Volume II on the British paper (drafting the opening frame letters around the same time?). Then, at the end of March and the beginning of April 1817, she revised Part [C] as Cr (and perhaps [B — C] as [B — Cr?]). However, this second hypothetical timetable, which would have involved a substantial amount of writing over what seems to be about twenty-one days in 1817, necessitates that the ordering evidence of the Clerval variants be discounted as accidental. Likewise the anomalous reference in Part D to Clerval's corpse looking like that of a young man of about twenty. And an extra chapter, for which we have absolutely no evidence, would need to be

inserted in Volume II of the Last Draft between the surviving opening fragment of Chapter 3 and the surviving closing fragment of "Chapter 6."

Both possible timetables have Mary essentially concluding the Last Draft with a revision of all, or just the last part, of the long "Chapter 4" that she had mentioned in her letter of 5 December 1816. The draft, or drafts, of that chapter are now lost except for the rewritten leaves that I have designated Part Cr. According to both timetables, the Cr rewrite would have taken place at, or around, the end of the creation of the Last Draft in 1817. But in the light of the second ordering timetable (the most probable in so far as it takes account of all the available evidence), we might now understand how it was possible that Frankenstein was, as M. K. Joseph supposes, essentially complete by late December 1816 and how it was, as Emily W. Sunstein determines, that Mary was working on a mid-point chapter around the same time. It is possible, indeed probable, that in December 1816 Mary was not faced with still having to write the last half of Frankenstein because that last half, Part D, was largely written before she wrote mid-point "Chapter 4."[4] In that case, what time Mary spent on the Last Draft from January to approximately 9 April 1817 was devoted to further revision, filling in, and patching.[28]

V

While there are only two extended analyses of the Frankenstein manuscripts (Murray; Mellor 57 — 68, 219 — 224) prior to the present one, analyses of the published text, whether the 1818 or the 1831 edition, or both, number in the hundreds. The question now arises, What are the interpretative implications of the analysis of the Last Draft that I have provided? The best way to answer this question is to review two articles which — amidst the sea of interpretation — are lent particular and quite startling support by what may

be deduced about Mary Shelley's inward-spiralling process of redrafting and revision: Marc A. Rubenstein's "'My Accursed Origin': The Search for the Mother in Frankenstein," and Joseph W. Lew's "the Deceptive Other: Mary Shelley's Critique of Orientalism in Frankenstein," an article that builds significantly on Rubenstein's.

Rubenstein argues that Frankenstein's concentric narrative arrangement implies a structural central point that is symbolically equivalent to the North Pole — the imagistic focus of Walton's quest and the approximate site of the novel's conclusion. That structural centre is the monster's rendition of Safie's account (in her letters to Felix) of her mother:

Safie's mother, Rubenstein declares, convincingly and for the first time, "is surely a cartoon, distorted but recognizable, of the author's mother, Mary Wollstonecraft" (169). The novel's hints at the Hyperborean myth of a warm polar womb are explained in terms of Mary's interest in her own conception and the tragic death of her mother following Mary's birth.

The account of Safie's mother occurs in what can be assumed to have been "Chapter 4" of Volume II of the Last Draft — that lengthy chapter towards which (according to my reconstruction of the drafting/copying/revision process) Mary had moved inwards in December 1816 (and in March 1817?); that chapter which seems to have given her particular difficulty to judge from the extant manuscript; and that chapter with which she essentially completed the drafting of Frankenstein. In other words, the theme that Rubenstein abstracts from the published text (with its spiral-like, concentric structure) — Mary's search for her own mother — was enacted by the inward-spiralling movement that characterizes the order in which she actually composed and revised her manuscript.

Although an analysis of "Safie and her unnamed father and mother, the only specifically Oriental personages within the narrative" (256), provides the climax of Joseph W. Lew's article, he devotes considerable space to Oriental elements in Frankenstein, as his projective synopsis (here abbreviated) makes clear:

Lew's argument is indebted to Edward Said, who describes the Orient as the source of one of Europe's "deepest and most recurring images of the Other" (1), and to Jacques Lacan. The Orient, the Other of Western civilization is — helped by puns on the French name for the "sea of ice" near Mont Blanc, Mer de Glace (sea mirror, mirror mother) — its self-reflection, its mirror image. Thus in the figures of Safie and her unnamed parents, "Having peeled back layer after layer, we find only the bourgeois nuclear family in Oriental drag" (282 — 283). Lew has moved on from the identification of Safie's mother with Mary's mother to the identification of Safie's father with Mary's father, William Godwin: Lew's entire argument, with its Lacanian conclusion that Frankenstein is "a hall of mirrors, a cautionary tale of a culture trapped in the mirror stage, and of the destruction that results from a way of seeing which can create only projections of Self" (283), interestingly complements mine in Frankenstein's Creation. In pursuing a broadly philosophical reading of Frankenstein, I distinguish between attempts to establish a relationship with, variously, a natural, a human, an inhuman and unnatural, and a transcendent Other, while exploring the sceptical, solipsistic undermining of all these forms of Otherness, all of which are exemplified by the monster. Lew's deceptive Oriental otherness might also be linked with this collapsing "four-fold Otherness" (Ketterer 105).[29]

Lew has not exhausted the Oriental aspects of Frankenstein.[30] Although

he makes a careful case for Frankenstein as a response to Percy Shelley's Orientalist poem "Alastor," he only once briefly notes (256) another of Shelley's Oriental poems, Laon and Cythna: or The Revolution of the Golden City, the epic he began while Mary was working on Frankenstein.[31] Shelley's Godwinian-inspired topic is a very idealized version of the French Revolution transposed to an Oriental setting. The temporarily successful revolt, organized by a brother and sister, Laon and the revolutionary feminist Cythna (who are loosely based on the atheistically-tarred Percy and Mary crossed with her mother) is celebrated by their incestuous union. In the censored version of the poem that was published as The Revolt of Islam (1818), Laon and the orphan Cythna are merely friends, much as Mary toned down the brother-sister relationship between Elizabeth and Victor Frankenstein in the 1831 edition.[32] As for the changed title of Shelley's poem, William St Clair explains that just as "a war to drive the Turks out of Europe was a universally accepted convention of a just revolution," so "Under another old convention writers could more safely attack religion if they called it Mohammedanism. Before the French Revolution philosophers teased the censors by pretending that their books were printed in Constantinople" (432). The same conventions should be applied to the story of Safie, who "arrived from Constantinople" (p. 57; Rieger 118.8), and her Turkish father who "had [li cancelled] inhabited Paris for many years when his person became [for some cancelled] [next insert in Percy's hand] ˄for some reason which I [the monster] could not learn˄ obnoxious to the government" (p. 57; cf. Rieger 118.4 — 6) and was condemned to death; after his escape he hires a vessel to take him "to [Can cancelled] Constantinople" (p. 59 [insert passage "64" to follow page 63]; Rieger 122.2). The sympathetic treatment here, combined with the Oriental subject matter, surely goes some way to explaining why, in her letter of 5 December 1816, Mary expressed the belief that Percy "would like" her "very long" "4 Chap. of Frankenstein" (I: 22).

In Laon and Cythna the tyrants quickly recover power and the eponymous protagonists suffer the fate that Mary's monster projects for himself: they are burnt to death. In the concluding canto, however, having been resurrected, they travel by canoe with their child to "Elysian islands, bright and fortunate, / Calm dwellings of the free and happy dead" (Complete Poetical Works 262; XII.31, 4727 — 28). (Compare the discarded Elysian framework of Mary's 1819 draft, "The Fields of Fancy," a work which also features incest; it was retitled Mathilda but not published until 1959. Similarly, Lord Verney, the eponymous character who stands in for Mary in her 1826 apocalyptic novel, The Last Man, hopes in conclusion "to moor my worn skiff in a creek, shaded by spicy groves of the odorous islands of the far Indian ocean" [342].)

The "Elysian islands" which surround the Temple of the Spirit are introduced in the symbolic opening canto of Laon and Cythna. The Poet-narrator and his Lady (and the wounded serpent enfolded in her breast) are in a bark:

This is the same Hyperborea that figures in the opening of Frankenstein: the North Pole presents itself to Walton's imagination as a "region of beauty and delight" where "the sun is forever visible" (Rieger 9.16 — 17).[33]

Of course, the Chinese-box structure of Frankenstein recalls what might be viewed as an Orientalist literary model — the framed story sequence of a work like the Arabian Nights' Entertainments. As it happens, Frankenstein

makes a direct reference to an episode from that work when he describes his reaction to discovering the secret of life: "I was like the Arabian, who had been buried with the dead, and found a passage to life aided only by one glimmering and seemingly ineffectual light" (p. 66; cf. Rieger 48.5 — 8). The "Arabian" is Sinbad, and the specific allusion is to his Fourth Voyage. But in characterizing him almost anonymously as "the Arabian," Mary Shelley surely is deliberately anticipating the only figure in her novel who is similarly identified (see Rieger 122.21) — Safie.

What finally needs to be added to Lew's account can only be supplied by the Last Draft of Frankenstein. As insightful as his article is, it does contain at least one significant blind spot, something he could have avoided had he had access to that Last Draft. After reviewing the case for paralleling Mary Shelley and "Walton's silent sister, the mysterious Margaret Saville [my emphases]," Lew points out that "in French . . . 'Saville' is almost homophonous with 'Safie.' Mary Shelley, Margaret Saville, and Safie become images of one another" (282). However, the name variants in the Last Draft, Amina/Maimouna/Safie, indicate quite clearly that it is Arabic pronunciation, not French, that is primarily relevant. All three names are associated with the Prophet Muhammad.

Āmina bint-Wahab was the mother of Muhammad; she died when he was six. Her first name is equivalent to āmina, meaning "peaceful or feeling safe," and to amina meaning "to be or feel safe." The differently pronounced Amīna (most famously the name of Solomon's wife) is the feminine of amīn meaning "honest or trustworthy," and derives from amuna meaning "to be reliable or faithful" (Baker 361). Maimouna is a transliteration of Maymŭnah, the eleventh or thirteenth (if two concubines are counted) and last surviving wife of Muhammad; she was a fifty-one-year-old widow in Mecca when Muhummad married her. Her name, bestowed on her by Muhammad and related to āmina and amīna, and to amuna and ma'mūn (meaning "reliable, trustworthy" [Baker 374]), means "auspicious" or "blessed."[34] Safie is a transliteration of Ṣafiyyah, the ninth wife of Muhammad (whose name means "praiseworthy, possessing fine qualities" [Baker 375]). Ṣafiyyah, who married him when she was eighteen (Mary was only one year older when she married Percy), was the daughter of the chief of the Jewish tribe of Bann Nadir, one of the Prophet's bitterest enemies; she was captured and converted to Islam. In describing "the works of the orientalists," which Frankenstein followed Clerval in reading, Frankenstein mentions "the smiles & [tears cancelled] ˄frowns˄ of a fair enemy" (Volume I, 99; Rieger 64.28) as a typical ingredient. Clearly, Ṣafiyyah was an historical example of such a "fair enemy" and it should be noted that Safie's complexion is "wondrously fair" (in the missing

pages 22 — 56 of Volume I of the Last Draft; Rieger 112.11).[35] Whether this imagistic connection is accidental or not, the Christian/Muslim lineage of Mary's character makes the name of the Jewish/Muslim Safiyyah/Safie a particularly appropriate choice among the names of those women linked with Muhammad. Safiyyah is Arabic for "confidante" or "bosom friend" and incorporates safa, "to be pure" (Baker 380). Lew, like presumably all readers of Frankenstein to date, assumed that "Safie" should be pronounced "Saf-ee," whereas in fact it might well be pronounced "Saf-i-yyah."

If Safie's mother and father correspond to Mary Wollstonecraft and William Godwin, and Safie to Mary herself, then Safie's lover, Felix/Muhammad, must correspond to Percy. Apparently Mary viewed Percy as a type of the Prophet. It is of interest to note that Mary read Voltaire's play Mahomet, ou le Fanatisme (1742) on 5 June 1818 (Journals I: 212), and that, in a letter dated 9 August 1830 to the publisher John Murray venturing various topics for his Family Library series, she writes: "A Friend suggested the life of Mahomet" (Letters II: 113). In the absence of that life by Mary, a reader interested in learning more about Āmina, Maymŭnah, Ṣafiyyah, and Muhummad, might well consult Martin Ling's Muhammad (1983), supplemented by al-Shati' Bint's The Wives of the Prophet (1971), and M. H. Zaid's Mothers of the Faithful (1935).

One might still wonder why, of all the names of all Muhammad's wives, Mary should have chosen Amina, Maimouna, and Safie. The answer would appear to be that the names occur (with two slight variations) in a narrative entitled "The Story of the three Calenders, Sons of Kings; and the five Ladies of Bagdad" (Weber I: 32 — 68). This tale sequence is part of "The Arabian Nights' Entertainments" which occupies most of Volume I of Henry Weber's magisterial 1812 compilation Tales of the East, a title that figures in Mary's reading list for 1815 (Journals I: 92). Volume I concludes with part of "The New Arabian Nights" (the "New Arabian Nights" that Mary records in her 1815 list was apparently the separate edition — with four fewer tales — identified by editors Feldman and Scott-Kilvert as Arabian Tales; or, A Continuation of the Arabian Nights Entertainments, a 1794 translation of a 1793 French translation [Journals I: 88]). Volume II continues with the remainder of "The New Arabian Nights," "The Persian Tales," "Persian Tales of Inatulla of Delhi," and "Oriental Tales." The concluding third volume contains "Mogul Tales," "Turkish Tales," "Tartarian Tales," "Chinese Tales," "Tales of the Genii," and "The History of Abdulla, the Son of Hanif."

The five ladies of Bagdad in "The Story of the three Calenders" (who are the ladies' guests) are sisters; they share the same father but three of them had one mother and the remaining two another. No doubt this narrative made an impression on Mary because of the parallels with her own family situation. Godwin had two wives and five children; he was father to two of them and stepfather to the other three. In the story sequence, of the three sisters

born to one of the mothers, one is named Zobeide and the other two (who as punishment for being jealous of Zobeide's projected marriage to a prince have been transformed into dogs — two black bitches) are unnamed (like the monster in Frankenstein). The jealousy of the black bitches suggests a parallel with Claire Clairmont's jealousy of Mary's relationship with Percy. The two sisters born to the other mother are named Amine and Safie. In an embedded story, a princess is "possessed by Genie Maimoun, son of Dimdim, who is in love with her" (Weber I: 47). Maimoun is the masculine form of Maimouna or Maymŭnah. Mary would have been struck by the name Maimoun because of its English language homonym: May moon. "Maië" was one of Percy's affectionate nicknames for Mary (Journals I: 80; Sunstein 96), and the moon (which is associated with the monster in Frankenstein) an agreed-upon symbol for her (Journals II: 434 n2; Dunn 226 — 227, 261, 274; Sunstein 194).[36] Thus the choice of the name Maimouna further corroborates the case for identifying Safie with Mary. The intricately interwoven narrative sequence includes six separately titled stories: "The History of the First Calender, a King's Son," "The Story of the Second Calender, a King's Son," "The Story of the envious Man, and of him that he envied" (this being a story within the Second Calender's story which underlines the envy/jealousy theme and includes the three page 47 references to Maimoun), "The Story of Zobeide," and "The Story of Amine." It might well have been as a result of the impact of the last story that Mary first alighted on the name Amina. Appropriately enough, Safie's tale is not related in "The Story of the three Calenders."

The second volume of Tales of the East includes "The History of the Birth of Mahomet" (Weber II: 616 — 631), among the "Oriental Tales." In this four-tale sequence, the yet-to-be-born Mahomet selects one Zesbet to be his mother and renames her "Aminta" (Weber II: 630), a corruption of Amina. A daughter takes on the new identity of a mother; and Mahomet, as the source of the name of that new identity, seeks to incestuously sire himself (his "real" father in fact died before he was born). To the extent that the monster is Frankenstein's double, Frankenstein too might be said to incestuously sire himself. The name "Zesbet" has elements in common with the name "Elizabeth" and may have been a source. Whether it was or not, the relationship between Zesbet and Aminta, like that between Elizabeth and Amina/Safie (or that between Elizabeth and Frankenstein's mother) is that of daughter and mother. The alternates Maimouna and Safie (Safiyyah) add the role of wife. To complete the gamut of archetypal female roles, it is only necessary to evoke that of the whore. This Mary Shelley does with her sometime choice (previously noted) of the name "Myrtella" for Elizabeth. As my appendix below reveals, "Myrtella" derives from the name of a mistress and courtesan. These archetypal identities are all dependent on the

complex procreative, erotic, and/or loving possibilities of sex. In usurping the procreative aspect of sex (thereby freeing the female from biological role determinism?), Frankenstein tragically (and unnecessarily?) denies the erotic and the loving aspects.

This analysis of altered character names (among other matters) has taken us a considerable distance into new areas of fact and of interpretation. Clearly, the Last Draft — and Fair Copy fragments — of Frankenstein have much to teach us about that novel's composition and interpretation beyond the matter of Percy Shelley's input. In fact, few literary manuscripts are quite so revealing. A new stage of Frankenstein scholarship will be initiated by their study.

Appendix: Five Other Altered Character Names in the Last Draft

Between Frankenstein's departure for the university of Ingolstadt and the arrival of Safie there are five variant character names in the Last Draft. They are reviewed here in the order of their appearance. Early in Part A of the Last Draft, the professor named Waldman first appears as "M. W. ------˄aldham˄" (p. 56; cf. Rieger 41.11); thereafter "Mr. Waldham" makes three appearances (twice on p. 57 [cf. Rieger 41.32, 41.35] and on p. 59 [cf. Rieger 43.12]). "Waldman" replaces "Waldham" in the following Chapter 5 of the Last Draft (p. 61; Rieger 45.9). The one subsequent reference, which is in Percy's hand, reverts to "Waldham": "M. Waldham expressed the most heartfelt exultation in my progress" (p. 62; cf. Rieger 45.26 — 27). Mary's alteration here was presumably made in the interest of verisimilitude — the German suffix "man" is more appropriate for the name of a professor at the German University of Ingolstadt than is the English suffix "ham."

The servant Justine is surnamed Moritz in the 1818 edition. But two cancellations in the Last Draft indicate that Mary originally gave her the too-English (or too-French) name "Martin" and later changed it to the German name Moritz. "Do you not remember [Ju cancelled] Justine [Martin? cancelled] ˄Moritz˄?" (p. 90; cf. Rieger 60.8), Elizabeth asks Frankenstein in a letter, before going on to recount Justine's history including "the death of [her father] M. [Martin cancelled] ˄Moritz˄" (p. 90; Rieger 60.11). Shortly afterwards there is a reference to "Mad[.] Martin" (p. 93; cf. Rieger 61.28 — 29: "Madame Moritz") but here "Martin" has not been deleted and "Moritz" has not been inserted.

As for a likely source, Emily Sunstein notes, "The names Moritz and Krempe, Victor's professors, and Victor's subsequent trip to England are drawn from the travel book by Carl Moritz that she had previously read" (340 n33). Mary's reading list for 1816 includes "X Moritz' tour in England," her "X" indicating that Percy also read the book (Journals I: 93). She is referring to Reisen eines Deutschen in England im Jahr 1782 (1783) by Carl Philipp Moritz, translated "by a lady" as Travels, Chiefly on Foot, through Several Parts of England in 1782 (1795). The nearest name to Krempe in the book is a "Mr. Kampe," who runs a German educational academy (Moritz 77).

Near the conclusion of Elizabeth's Justine letter occurs the following concentrated passage with its two name changes:

The last of these names may be traced to Mary's biography. While at the Maison Chapuis, Mary and Percy hired a Swiss nursemaid named Louise Duvillard, whom they called Elisa (see Sunstein, "Louise Duvillard" 27 — 30; Letters I: 36, 37 n9; and Journals I: 201 n1, where an "i" is [mistakenly?] added to the surname of her parents: "M. and Mme Jean Duvilliard"). Her first name, "Louise," seems to have simultaneously migrated to the immediately preceding reference to little William's favourite girlfriend "Louisa [is cancelled] [tab cancelled] Biron" (p. 93; cf. Rieger 62.5) and to the immediately following reference to "Louis Manoir": "Your favourite schoolfellow Louis Manoir [ma cancelled] has suffered several misfortunes [d cancelled] since the departure of Clerval from Geneva [b (?) cancelled][.] But he has [atready cancelled] already recovered his spirits, and he [is] reported to [m cancelled] be on the point of marrying a very lively pretty french woman — Mad. Tavernier" (p. 94; cf. Rieger 62.12 — 16). A Tavernier, an agent of Percy's friend, Thomas Hookham, according to R. Glynn Grylls' pioneering biography (33 n1), was a Paris banker or money lender who lent Shelley £60 see (Journals I: 9 and 9 n3, 10 — 11, 28; and Sunstein 85).

The most extraordinary altered name also occurs in Part A of the Last Draft. Regarding the accusation of murder levelled at Justine, Frankenstein's remaining brother, Ernest, observes that "[Myrtella will cancelled] Elizabeth will [the replacement appears in the margin] not be convinced notwithstanding all the evidence" (p. 117; cf. Rieger 74.16 — 17). The strange alternate name for Frankenstein's betrothed occurs on only one other occasion in the Last Draft — again in Part A: "At eight in the evening we arrived at Chamounix. [Myrtella cancelled] ˄My father & Elizabeth were˄ very fatigued" (p. 148; cf. Rieger 90.28 — 29). It seems clear that on both occasions the alternate name was quickly cancelled. Lemprière's Classical Dictionary of Proper Names describes a "Myrtăle" as "a courtesan of Rome, mistress to the poet Horace," and gives Horace's Odes I, number 33 ("Albi, ne doleas"), as the only source (394). Apparently the name Myrtăle, "derived from the myrtle's association with Aphrodite", was "often borne by freedwomen" (Nisbet and Hubbard 375).

Ode I.33 consoles the poet Albius Tibyllus on the loss of a mistress who had deserted him; in its fourth and last verse Horace proffers the experience of his own recovery from a similar loss:

In this translation, Lord Dunsany has anglicized the Latin "Myrtăle" by clipping off the final "e". In her journal Mary includes Horace's Odes in her reading list for 1816, and specifically mentions reading them in Bath on 3 and 5 December 1816 (I: 97; I: 148, 149); that is to say she was reading the Odes while writing the Last Draft of Frankenstein. But why — since "Myrtella" clearly is Mary's phonetic equivalent of "Myrtăle" — was she so struck by this single reference to Horace's ex-mistress? Did she in some way relate the relationship between Horace and Myrtăle to what she might have seen as the precarious relationship between Percy and herself (and hence to that between Frankenstein and Elizabeth)? At the same time, various aspects of Horace's personality might have seemed relevant to what Mary saw herself advocating in Frankenstein: Horace's enthusiasm as a young man for republican ideas and the delight he expresses in his poetry for the simple life, the beauties of nature, and the joys of friendship.

The two surprising appearances of "Myrtella" recall the two equally bewildering occurrences of both "Carignan" and "Amina." Was Mary once more absent-mindedly copying a superseded name from a preliminary rough draft? Was there, then, a version of Frankenstein in which Frankenstein's friend was called Carignan, the cousin he married Myrtella, and Felix's fiancée Amina?

Works Cited

• Arabian Tales; or, A Continuation of the Arabian Nights Entertainments . . . Newly Translated from the Original Arabic into French by Dom Chaves [Denis Chavis], a Native Arab, and M. Cazotte, Member of the Academy of Dijon. 3 vols. London: Faulder, Hookham & Carpenter, etc., 1794.
• Baker, Mona. "Supplement I: Common Names in the Arab World." In A Dictionary of First Names (New York: Oxford University Press, 1990): 351 — 386.
• Barker-Benfield, Bruce C. "The Frankenstein Notebooks: Technical Notes, Watermarks, Collations." Unpublished 21 page report dated 7 February 1995.
• Bint, al-Shati'. The Wives of the Prophet. Trans. Matti Moosa and D. Nicholas Ranson. Lahore, Pakistan: Sh. Muhammad Ashraf, 1971.
• Clubbe, John. "Mary Shelley as Autobiographer: The Evidence of the 1831 Introduction." The Wordsworth Circle 12 (Spring 1981): 102 — 106.
• The Collected Works of Horace. Trans. Lord Dunsany and Michael Oakley. London: J. M. Dent & Sons Ltd., 1961.
• The Complete Poetical Works of Percy Bysshe Shelley. Vol. II (of 4), 1814 — 1817. Ed. Neville Rogers. Oxford: Clarendon Press, 1975.
• Crouch, Laura E. "Davy's A Discourse, Introductory to A Course of Lectures on Chemistry: A Possible Scientific Source of Frankenstein." Keats-Shelley Journal 27 (1977): 35 — 44.
• The Diary of Dr. John William Polidori, 1816. Ed. William Michael Rossetti. London: Elkin Mathews, 1911.
• Dunn, Jane. Moon in Eclipse: A Life of Mary Shelley. New York: St. Martin's Press, 1978.
• Grylls, R. Glynn. Mary Shelley: A Biography. London: Oxford University Press, 1938.
• Hammond, Ray. The Modern Frankenstein: Fiction Becomes Fact. Poole, Dorset: Blandford Press, 1986.
• Hunt, Marie-Hélène. Monstrous Imagination. Cambridge, Mass.: Harvard University Press, 1993.
• The Journals of Mary Shelley, 1814 — 1844. Ed. Paula R. Feldman and Diana Scott-Kilvert. 2 vols. Oxford: Clarendon Press, 1987.
• Joseph M. K. "Appendix A. The Composition of Frankenstein." In Mary Shelley, Frankenstein; or, The Modern Prometheus (Oxford: Oxford University Press, 1969, 1980): 224 — 227.
• Ketterer, David. Frankenstein's Creation: The Book, the Monster, and Human Reality. English Literary Studies Monograph Series, no. 16. British Columbia: University of Victoria, 1979.
• Leask, Nigel. British Romantic Writers and the East: Anxieties of Empire. Cambridge: Cambridge University Press, 1992.
• Lemprière's Classical Dictionary of Proper Names mentioned in Ancient Authors Writ Large. Third Edition. London: Routledge & Kegan Paul, 1984. The French original was first published in 1788.
• The Letters of Mary Wollstonecraft Shelley. Ed. Betty T. Bennett. 3 vols. Baltimore: The Johns Hopkins University Press, 1980, 1983, 1988.
• Lew, Joseph W. "The Deceptive Other: Mary Shelley's Critique of Orientalism in Frankenstein." Studies in Romanticism 30 (Summer 1991): 255 — 283.
• Ling, Martin. Muhammad: His Life Based on the Earliest Sources. Cambridge: The Islamic Texts Society, 1983.
• Majeed, Javel. Ungoverned Imaginings: James Mill's The History of British India and Orientalism. Oxford: Clarendon Press, 1992.
• Mellor, Anne K. Mary Shelley: Her Life, Her Fiction, Her Monsters. London: Methuen, 1988.
• Moore, Thomas. Letters and Journals of Lord Byron: With Notes of his Life. 2 vols. London: John Murray, 1830.
• Moritz, Carl Philipp. Travels of Carl Philipp Moritz in England in 1792. A reprint of the English Translation of 1795. London: Humphrey Milford, 1924.

---

276




• Murray, E. B. "Shelley's Contribution to Mary's Frankenstein." Keats-Shelley Memorial Bulletin 29 (1978): 50 — 68.
• Nisbet, R. G. M., and Margaret Hubbard. A Commentary on Horace: Odes Book I. Oxford: Clarendon Press, 1970.
• Reiman, Donald. "The Composition and Publication of The Revolt of Islam." In Shelley and His Circle, ed. Kenneth Neill Cameron and Donald H. Reiman (Cambridge, Mass.: Harvard University Press, 1961): 145 — 155.
• Rieger, James. "Dr. Polidori and the Genesis of Frankenstein." Studies in English Literature 3 (1963): 461 — 472. Reprinted in slightly revised form in Rieger, The Mutiny Within: The Heresies of Percy Bysshe Shelley (New York: Braziller, 1967): 237 — 247. Cited as "Rieger, 'Dr. Polidori.'" For the citation "Rieger," see Shelley, Mary Wollstonecraft, Frankenstein.
• Rostaing, C. H. Dictionnaire étymologique des noms de lieux en France. Paris: Libraire Guénégaud, 1963.
• Rubenstein, Marc A. "'My Accursed Origin': The Search for the Mother in Frankenstein." Studies in Romanticism 15 (Spring 1976): 165 — 194.
• Said, Edward W. Orientalism. New York: Pantheon Books, 1978.
• Shelley and Mary. Prepared for the press by Jane, Lady Shelley. 4 vols. 1882. (Twelve copies printed according to Grylls vi; copy consulted: that of Percy Shelley's biographer, Edward Dowden, with his annotations, in the British Library [Ashley 4088].)
• Shelley, Mary Wollstonecraft. Collected Tales and Stories. Ed. Charles E. Robinson. Baltimore: The Johns Hopkins University Press, 1976.
• [_______.] Frankenstein: or, The Modern Prometheus. 3 vols. London: Lackington, Hughes, Harding, Mavor, & Jones, 1818.
• _______. Frankenstein or The Modern Prometheus (The 1818 Text). [Including the 1831 Introduction.] Ed. James Rieger. Indianapolis: The Bobbs-Merrill Company, Inc., 1974. (Reprinted by The University of Chicago Press, 1982.) Cited as "Rieger."
• _______. The Last Man. Ed. Hugh J. Luke, Jr. Lincoln: University of Nebraska Press, 1965. Originally: London: Henry Colburn, 1826.
• _______. Mathilda. Ed. Elizabeth Nitchie. Extra Series #3 of Studies in Philology. Chapel Hill: University of North Carolina Press, 1959.
• _______. The Rough Draft and the Fair Copy of Frankenstein by Mary Shelley. Dep.c.477/1, and Dep.c.534/1 and /2; The Abinger Deposit: Papers of P. B. Shelley, W. Godwin, and Their Circles; The Bodleian Library, Oxford.
• _______, with Percy Shelley. History of a Six Weeks' Tour through a Part of France, Switzerland, Germany, and Holland: with Letters Descriptive of a Sail round the Lake of Geneva, and the Glaciers of Chamouni. London: T. Hookham, Jun.; and C. and J. Ollier, 1817.
• St Clair, William. The Godwins and the Shelleys: The Biography of a Family. London: Faber and Faber, 1989.
• Sunstein, Emily W. "Louise Duvillard of Geneva, the Shelley's Nursemaid." Keats-Shelley Journal 29 (1980): 27 — 30.
• _______. Mary Shelley: Romance and Reality. Boston: Little, 1989.
• Walling, William A. Mary Shelley. New York: Twayne, 1972.
• Weber, Henry William, ed. Tales of the East: Comprising the most popular Romances of Oriental origin; and the best imitations by European Authors: with new translations, and additional tales, never before published; To which is prefixed an Introductory Dissertation, Containing an Account of each work, and of its author, or translator, By Henry Weber, Esq. 3 vols. Edinburgh: John Ballantyne and Company, 1812.
• Wolf, Leonard. The Annotated Frankenstein. New York: Clarkson N. Potter, Inc., 1977. (Revised as The Essential Frankenstein [New York: Plume, 1993]; my references are to the differently paginated first version.)
• Zaid, M. H. Mothers of the Faithful. Calcutta: Art Press, 1935.
• Zonana, Joyce. "'They Will Prove the Truth of My Tale': Safie's Letters as the Feminist Core of Mary Shelley's Frankenstein." Journal of Narrative Theory 21 (Spring 1991): 170 — 184.

Notes