University of Virginia Library

Sons and Lovers: The Surviving Manuscripts From Three Drafts Dated by Paper Analysis
by
Helen Baron

Sons and Lovers is the autobiographical novel of an author whose biography has, since his death, rivalled his art in commanding the attention of readers and critics. The two years he spent writing his four drafts of the novel (October 1910 to November 1912) witnessed the most momentous events in his life: his mother ("my first, great love")[1] died, he left his teaching job for the risky experiment of living by his pen, he went to live out of England for the first time, and he began sharing his life with Frieda Weekley; he also brought to an end his long relationship with Jessie Chambers, and was engaged for fourteen months to Louie Burrows. The four drafts that he wrote in those two years bore witness to his efforts progressively to understand and face or evade the implications of those events while living through them.

Lawrence himself called the work "autobiography" (Letters, I, 490), but it was no simple record; by writing it he was examining a central life-problem that was Oedipal in character, and doing so not only before Freud's theories had made their impact on Europe, but also through the medium of the novel

rather than the treatise or essay. His method of arriving at his conclusions was to renew his efforts at interpretation with each successive draft. Thus the sequence of drafts would show the steps by which he made progress in the uncharted territory of psychological self-analysis, and how he learned to reconceive the potentialities in the novel form for the purpose of such fictional-biographical exploration. The full drafts would therefore show more than developments in style, they would reveal how Lawrence brought into shape an innovative work of art; in addition, they would document his oscillations between one interpretation and another as he pondered the conflicting pressures in his personal history, and they would provide some insights into the emergence of his verdicts.

Echoes of a conflict between rival viewpoints linger on into the published novel in the recurrent tones of judgement, accusation and self-justification. This tendency of the novel to sound like a debate has been reinforced from without by the existence of Jessie Chambers's rival interpretation in her memoir, written after Lawrence's death.[2]

All these complexities cause the surviving early drafts to be the subject of a more intense and varied curiosity on the part of readers of the novel than many an author's "foul sheets" would necessarily be. But because each draft is so intimately bound up with Lawrence's life in those two years, it is not enough simply to read and compare different versions; we need to know as closely as possible when the surviving pages were written.

Alas, not everything has survived, but in addition to the complete final manuscript miscellaneous portions from previous drafts are extant. No serious attempt has been made to date these and discover how they all relate to each other.

The University of California published in 1977 a beautiful photographic facsimile of the Sons and Lovers manuscripts acquired in 1963 and held at the Bancroft Library, Berkeley, which include not only the final manuscript itself but some rejected sheets from earlier drafts.[3] The editor, Professor Schorer, labelled these sheets "Fragments" and divided them up at follows:

           

Fragment 1	Chapter I	pp.1-8
2		pp.26-36
3	Chapter II	pp.44-45,46a-49a,46-58
4		pp.68-69
5	Chapter III	pp.68-82
6	Chapter IV	pp.83-85

Fragment 6 is in fact a continuation of Fragment 5: no pages are missing between the end of chapter III on page 82 and the beginning of chapter IV on page 83. Professor Schorer registered his surprise (p.[3]) that Fragment 4 contained page-numbers also found in Fragment 5: the answer to this enigma is that Fragment 4 belongs to a different period of composition from all the other Fragments except the inserted pages 45, 46a-49a of Fragment 3.

Any reader of the facsimile of the final manuscript itself is bound to notice that it contains long and short sequences of varied paper, some sections of which have had their page-numbers altered once or even twice. Did these renumbered sections originally belong to earlier drafts? If so, which drafts? When were those passages written? How do they relate to the other, discarded, portions of drafts: the Fragments at Berkeley and the manuscripts now held at the Humanities Research Center of the University of Texas at Austin?

It is strange that Professor Schorer made no mention of the latter, for they had been described in print (albeit not wholly accurately) by Harry T. Moore;[4] and they were the only Sons and Lovers manuscripts listed in the edition available to him of Warren Roberts's scholarly and authoritative A Bibliography of D. H. Lawrence.[5]

The University of Texas H.R.C. has held since the early 1960's a 271-page manuscript of an early draft which is there labelled The Paul Morel Manuscript; and in November 1980 the Center also acquired a twenty-three-page section from a draft written in Lawrence's hand but annotated with crititisms and revisions by Jessie Chambers.

It is well known that Lawrence tried to improve his progress with the writing of this novel by involving in the process of composition at several stages one of the main participants in the life-history recounted therein, Jessie Chambers. As a result she wrote for him biographical notes, and critiques of his drafts, some of which have fortuitously survived. The existence of her marginalia shows that she also wrote her objections and criticisms on the manuscript of the novel itself, and the result is a richly suggestive document. The acquisition of this manuscript completes the H.R.C.'s holding of the small selection of Jessie's written contributions to Sons and Lovers known to have survived. The others are three episodes and two commentaries, making six manuscript sections in all. It has never been satisfactorily established how these Jessie Chambers manuscripts (labelled the 'Miriam Papers' by Harry T. Moore[6]) relate either to each other or to the successive drafts of Sons and Lovers.

When my co-editor and I were pondering our microfilms of the surviving

manuscripts in Cambridge, England, in preparation for a study of the originals in Berkeley and Austin, and wondering whether it would be possible to settle some of these unanswered questions, the significance dawned upon me of a remark in one of Lawrence's letters: "The 112 pages of Paul are pages such as this on which I write. Am I a newspaper printing machine to turn out a hundred sheets in half an hour?" Letters I, no. 264. Did Lawrence's expostulation contain a clue as to where the key to this jigsaw-puzzle might be found?

That letter, written on 9 May 1911 to Louie Burrows and now held at Nottingham University Library, was written on a 21-lined paper which had been torn off a pad with the aid of such large perforations that the sheet had acquired a markedly "serrated" top edge. These same features, 21 lines and a serrated top edge, appeared to be present also (as far as one could tell from a microfilm) in the first surviving pages, 72-165, of The Paul Morel Manuscript. Were some of these the pages Lawrence had been referring to in that letter? Did he regularly write letters on the same paper that he was using for his novel (and stories and poems)? If so, the letters, as physical documents, would provide a dated framework for reconstructing the history of the layered manuscripts and miscellaneous fragments that have survived.

Preliminary study of the holdings in the library of Nottingham University gave me reason to believe that some correlations would be revealed if we undertook as full a survey as possible of the relevant documents. Nearly all the extant letters written between October 1910 and November 1912 are now held in the Nottingham University and County Libraries, the H.R.C. at Austin and the Berg Collection of the New York Public Library. Some comparison of manuscripts of the novel with letters could be made physically at Austin, but for identification of paper-types between the various library holdings some form of sufficient paper-description needed to be devised.

The comprehensiveness of James T. Boulton's The Letters of D. H. Lawrence Vol. I 1901-1913 permitted us the great convenience of referring to each letter by the enumeration in that volume (which also gave economical identifications between letters written on the same day and even to the same recipient). In the larger period September 1910 to May 1913, that we chose prudentially to cover, there are 402 items in Letters I (Nos 178-579). Of these, 83 are postcards and one was written on a concert programme; 6 survive only in typed copies and 27 as quotations in books, newspapers or private communications; a further 9 letters are in private hands; 7 are at UCLA[7] and one at the University of Indiana.[8] The remaining 268 original letters we examined

and described for comparison with the paper-types found in the 882 surviving leaves of manuscripts of the novel.

In this article I report the results of the research, beginning with an account of the problems encountered, the methods evolved and the picture that emerged of Lawrence's habits in the use of stationary. The grid of descriptive questions we assembled reflects the range of paper-types being covered, and the terminology we settled for reflects the practical rather than theoretical emphasis of the task. Scholars engaged on similar projects, whether in relation to further manuscripts of D. H. Lawrence or to other authors, may perhaps benefit from the information and adapt the techniques described here.

Then, although a wide range of results flows from the precision in dating made possible by the paper research, this article is confined to explaining the chronology of composition and interrelations between the surviving portions of drafts in terms of paper-batches. This information is summarised in two tables at the end: they provide the essential factual foundations for readers of the California facsimile who wish to develop their observations about Lawrence's revisions.

Lawrence's statements in his letters give an outline composition history of Sons and Lovers as follows. It was written, in effect, in four stages: Lawrence wrote four drafts starting from the beginning each time and he conducted a separate, thorough revision of one draft; but he did not complete all these four stages. He wrote Stage I (unfinished) from October to December 1910; Stage II (unfinished) from 13 March to July or October 1911; Stage III he began on 3 November 1911 and continued until perhaps 19 November when he fell ill, and then after convalescing he began again sometime between 9 and 23 February 1912 and finished on 11 April 1912; on 3 May he took this IIIa manuscript with him to Germany, and Stage IIIb was a revision of this version carried out from 16 May to between 2 and 9 June 1912; in September he arrived in northern Italy where he wrote Stage IV from 7 September to 19 November 1912. The paper research modifies some of this outline as well as providing dates within which each sequence of paper was used.

After the completion of the manuscript stages, the novel was further altered during the ensuing months. First, the publisher's reader at Duckworth's, Edward Garnett, reduced its length by a tenth by crossing out passages throughout the final manuscript. Secondly, Garnett sent the edited manuscript direct to the printer, Billing and Sons in Guildford, where it was used as copy: the compositors introduced a few substantive variants and thousands of changes in punctuation. Finally, two rounds of proofs were revised by Garnett and Lawrence before publication on 29 May 1913.

The change of title from 'Paul Morel' to Sons and Lovers during composition is not perfectly reflected now in the labels by which the surviving manuscripts are known. From his very first mention of the novel in October 1910, Lawrence called it Paul Morel, and it was not until mid-October 1912, when he had written three-fifths of the final manuscript and only a month before he completed and despatched it, that he used the title Sons and Lovers.

Therefore all the references in Lawrence's letters during those first two years, that is for almost all the manuscript stages, are to Paul Morel. But Paul Morel is the name now given at the H.R.C. to the 271-page manuscript (largely of Stage II) held there, and Professor Schorer (pp. [2]-[3]) suggested the title Paul Morel or PM3 for the Fragments at Berkeley. In this article Lawrence's uses of titles in letters will be quoted as they are printed in Letters I; 'Paul Morel' will be the form for referring to mention of the novel in the records for that period; The Paul Morel Manuscript will be used for the H.R.C. manuscript of Stage II; the "Final" manuscript for the complete manuscript of Stage IV, and "Fragments" for the discarded pages at Berkeley; the "section" of chapter IX for the twenty-three pages of a draft held at H.R.C.; the "remains" of Stages IIIa and b for the reconstruction of that draft from the surviving evidence; and "items" for the distinct manuscript sequences belonging to the so-called "Miriam Papers".

The Paper Research

Because we were concerned with distinguishing between a number of similar papers by minute differences in measurement, it was helpful to refer to them not by the trade names for the different paper-sizes but by their distinctive fine-measurements which we made entirely in millimetres. The height of each leaf needed to be measured at both side edges and the width at both top and bottom, because in some batches they varied, and if Lawrence reversed a page the variation could give a false impression of a different paper-batch. Many of the later papers were large sheets folded into quires, and because these had been torn unevenly down the fold it was necessary to reassemble conjugate leaves to make accurate measurements across the width.

We chose, in fact, to trace all the patterns of conjugate leaves and original quires, but these are not recorded in Table 1. They led us to discover that Lawrence, who wrote on one side of the paper only, left the quires intact and occasionally inserted leaves from earlier drafts among the pages of the quires. When Garnett edited the manuscript he crossed out text and wrote the delete symbol in the margin with the quires still intact, so that some of his writing which overflowed the fold is now on the reverse of conjugate leaves. It is most probable that the quires were torn in the printing shop for distribution to the compositors.

We found the use of a micrometer for measuring paper thickness (calibration) in hundredths of a millimetre very helpful; but measurements needed to be taken at various points on a page because some papers had uneven density.

Describing the colour of paper is problematic, not only because most papers are a variety of white, but also because they discolour with age differently according to the different conditions in which they have been kept by their various owners. Dr. Philip Gaskell of Trinity College, Cambridge, whom I consulted about this research, suggested we should assemble a collection of samples of papers of graduated colour shades and label them numerically. This excellent advice we did not have time to follow, and the

resulting colour descriptions we resorted to served only as subjective aidemémoire. In the end, colour difference did not constitute the sole basis for any identification—despite the one very distinctive blue paper, which also had very distinctive line-numbers—and I doubt if in advance we would have thought to extend the range of samples to cover the variety of flecked-brown papers Lawrence acquired by the Lago di Garda.

Many of Lawrence's papers, especially the early ones bought in England, were lined and we had to develop a number of definitions for distinguishing between them. The greatest problems presented by this enquiry concerned the very similar pads of uniform lined paper that Lawrence used while writing the now critically dispersed Stage IIIa draft. These papers, of 24 and 26 lines (see the e,g,k and h,j,l papers in Table 1), are easy to confuse; and to make matters worse he used further similar pads in his letters which do not occur in the surviving remains of the novel. Therefore, in addition to the routine measurements of spaces between lines (only the averages of which are given in Table 1), spaces above the top lines and below the bottom lines (which for convenience we called "top and bottom margins") and any side margins, we also paid attention to the direction of the slope with which the lines were characteristically printed on the page (or which characteristically resulted from the way the blocks were cut), to the quality of the printing, and whether the lines were printed differently on front and back of the paper.

As a result of my preliminary discussion with Dr. Gaskell, we decided upon five categories to define paper quality,[9] which turned out to be very useful. The first two categories were the characteristics the paper displayed by transmitted light, its "See-through features", consisting of transparency or density and the presence or absence of watermarks. Degrees of transparency were often significant, and although there were few watermarks they appeared and disappeared in clear-cut sequences in the letters, and this contributed useful dating evidence for the novel.

The other categories of paper quality were descriptions of the general impression the paper-finish displayed to the eye under normal and raking light. Under the heading "Finishing features" we recorded whether the paper was shiny or matt; whether it was flexible or brittle (including our impression of the noise a page made when shaken if this was a significant corroboration of its stiffness or suppleness); and its calendering. "Calendering" was not recorded when paper was "laid", that is, had chain-and-wire lines; most of the papers were "wove", but we doubted whether the labour of measuring the mesh would be rewarded with the clear distinctions we were looking for, and instead we recorded as an aide-mémoire our impression of the surface patterns under a raking light. These various subjective descriptions of finishing features were, like paper-colour, only corrective or supportive of any identification, made primarily on the basis of measurements, lines and watermarks in combination with idiosyncrasies such as the "serrated" top edge.

Finally, under the heading "Format" we recorded any evidence of the form the paper batch had taken whether pad (i.e. glued block), quire or sheet. We found that some single sheets showed evidence of being torn from a pad while others from the same pad had been torn off without any betraying mark being sustained.

When Lawrence went back to Croydon to continue teaching (after his mother's death on 9 December 1910) he wrote his letters on a wide range of papers. If he was writing from his lodgings in Colworth Road it included not only pages from the supply he was using for his novel or poems, but also mourning notepaper with black edges and Colworth Road headed notepaper. When he wrote letters from school he used pages torn from school registers or exercise books, or even Croydon Education Committee headed paper, and, once, an Accident Report Form. When he went to stay with people he used their headed notepaper; and the letter on a concert programme also comes from this period.

From 19 November 1911, when he fell ill, to 28 February 1912, there was a sudden change from this miscellaneity to a simple sequence of good-quality letter-writing paper. He started the first, watermarked "Court Royal", on 15 November and finished it on 13 December 1911; the second, watermarked "Classic Parchment", was then used equally consistently from 17 December 1911 to 7 January 1912. That last letter on Classic Parchment was the first he wrote from his place of convalescence, Compton House, Bournemouth; he then discovered the guest-house's supply of headed notepaper, and set his Classic Parchment aside to use up later (10 February to 17 April).

After his return to Eastwood on 9 February he was mobile again and inevitably his letter-paper became miscellaneous again (once the Classic Parchment was finished), and began to include the lined paper he was currently using for his novel and poems. Then when he went abroad in May 1912, his letter-paper became, if possible, even more varied; but it did include the two surviving papers introduced in Stage IIIb and four of the six new papers used for Stage IV.[10]

In the surviving manuscripts of Sons and Lovers there are nineteen different papers. These I have labelled in alphabetical order (omitting i and o) to correspond with the chronological order of their use in the manuscripts,

using capitals for the papers first introduced in Stage II (A-D), then lower case for those first used in Stage IIIa (e-l) and IIIb (m-n), and capitals again for Stage IV (P-U). Lower case is used for Stage III for instant recognition because of the difficulties presented by the dispersal of the manuscript remains of this draft.

The character of the sequence of papers in the manuscript drafts of the novel is from cheap-quality lined papers in England to quires of unlined paper in Italy, where Lawrence wrote much of the novel out afresh for Stage IV.

The information derived from the letters needs to be used with care for a number of reasons. Lawrence used several uniform pads of lined paper, and while sequences of similar paper distinguish themselves from each other by physical comparison in the body of The Paul Morel Manuscript or the Final manuscript and Fragments, transporting these fine distinctions for comparison with individual letters of one or more leaves in libraries many miles distant needs to be done with cautious scepticism to prevent misidentifications. Furthermore, Lawrence must often have had a variety of papers in use simultaneously and used them at intervals (as can be seen with the S paper in Table 1). On the other hand, some papers are very distinctive, and because they make their appearance in the letters for a short period and then disappear, they do provide real evidence about the novel. If this evidence is used with circumspection and in conjunction with the other information available, not only does it establish a number of facts, it is indeed the sole means of answering some questions.

Table 1

Table 1 takes each paper in chronological order of its use in the drafts of the novel, which is identical with the order in the letters except for paper S, and provides for each a description which is sufficiently full to help in the recognition of most papers.

Where it has been possible to identify any paper in other works the details are added below each description, but this question has been investigated only in the collections at Nottingham University Library; the Bancroft Library, Berkeley; H.R.C., Austin; and the Berg Collection at the New York Public Library. Finally, the letters written on the same paper are listed by their numbers in Letters I and their dates.

Table 2

Table 2 lists the surviving pages[11] of each draft of the novel, and shows the sequences of paper-types within each manuscript. The survivals from each stage of composition are in broad outline as follows: Stage I is lost; Stage II survives in large part, as 264 pages of The Paul Morel Manuscript, which

lacks both its opening and concluding pages. Some pages were transferred by Lawrence into subsequent stages of composition and 13 are now in the Final manuscript. Seven unnumbered pages of the novel's opening now kept with The Paul Morel Manuscript do not belong to this draft, but will be shown below to have been an abortive start to Stage IIIa. Of Stage III proper, both a and b, 161 pages (probably about one-third) are preserved piecemeal in a variety of ways. There are the 58 pages of Fragments rejected by Lawrence during the writing of Stage IV, now in the Bancroft Library; there are the 80 pages that Lawrence transferred a few at a time into Stage IV, now in the Final manuscript; and there is the section of 23 pages from chapter IX now at the H.R.C. Stage IV is the Final manuscript in the Bancroft Library.

Because Stages II, IIIa and IIIb no longer stand complete in their original form, part of the information in Table 2 is an abstract reconstruction. Many pages from these three periods of composition are now missing; and because those that survive are now dispersed in separate sections and in the Final manuscript, it is necessary to add to the brief information about each sequence of pages as they originally stood in each draft, a note of where those pages are now located.

The dates of each draft are given at the head, and the dates of the paper's use elsewhere are summarised in addition (but the paper's use elsewhere is omitted if precise dates cannot be given, as for pages of the Love Poems MSS).

The version of the novel for which this exercise of reconstruction is most crucial is, of course, Stage III: it would be impossible to detect that the material on m paper belonged to Stage IIIb and not IIIa except by investigating the letter-paper. Similarly it is only possible to deduce that Lawrence wrote pages 1-74 of Stage IIIa before his illness in November 1911 from the fact that the paper (f) was used up during the next month in almost every letter, and then a new supply of paper followed with the same consistency.

THE COMPOSITION STAGES

The description below of the composition of the second and third drafts of Sons and Lovers follows as exclusively as possible two threads of interest: the information derived from the letter-papers about the paper-batches in the surviving drafts, and the new discoveries about the composition history. These two strands have to be extricated from the complex weave of Lawrence's other writing, the pressures and influences exerted upon him, and his various friendships during this period, the whole pattern of which cannot be recounted here, but brief sketches of which are given when they impinge directly on the composition of the novel.

Stage II Lawrence's letter of 9 May 1911 (no. 264) containing the paper clue was one of eleven written between mid-March and the end of May in which he reported to Louie Burrows on his progress with the novel.

In the first, of 13 March (no. 236), he announced that he had begun 'Paul Morel' again and added that he feared it would be a "terrible" (i.e. a painful) novel. He made the same announcement to two other people in the next two days and then did not mention the novel again until 12 April (no. 254), by

which time he was beginning to worry about his rate of progress. He told Louie: "I have just done one folio, a dozen MSS pages, of Paul Morel. That great, terrible but unwritten novel, I am afraid it will die a mere conception." By "just" he seems to have meant "just now" and not "only". These folios must have been among the opening 71 pages which are now missing, for the first pages that survive (pp. 72-165) are the "serrated" paper, now labelled A paper, which Lawrence began using for his letters in his next letter to Louie of 24 April (no. 256). He did not use folios again until page 255.

The anxiety he expressed on 12 April deepened, and on 28 April (no. 259) he wrote to Louie in an outburst that could be read as an appeal for moral support: "Do you know I simply cannot work. I have done only about five pages of MSS, 'Paul Morel'; and that only from sheer pressure of duty. I don't want to work: and I don't care a damn about it." She must have responded to the appeal, for three days later Lawrence was evidently making use of her encouragement to force himself through these initial difficulties (no. 261): "At your behest I wrote yesterday fourteen pages of Paul Morel, and I sit with the paper before me to continue when this is done."

Then, three days after this, on 4 May (no. 262), he gave the first indication where he was in the text, and even recorded how he felt about what he had written so far: "I have written 90 pages of Paul Morel. I think about 7 of these pages may be called amusing, and 20 perhaps pleasant. The rest are "navrant" [i.e., distressing]. I wonder how Paul will work out."

Of these ninety pages, numbers 72-90 survive in The Paul Morel Manuscript. The first three were originally numbered 74, 75 and 76, and perhaps Lawrence renumbered them as he removed a few earlier pages.[12] The text on page 72 begins mid-sentence in the middle of a satirical scene in which Walter Morel, ill in bed, receives a visit from his boon companion, Jerry. The loyal friend has smuggled in for him a longed-for bottle of beer, and is moved to tears by the unwonted difficulty Walter has, because of his enfeebled condition, in swallowing more than a few grateful mouthfuls. This scene ends on page 77.

The rest of the chapter recounts Walter Morel's convalescence, his rejection by the baby Paul, and Paul's casting aside by his mother into the hands of a nurse at the age of seventeen months because of the birth of Arthur. At this last event, Walter Morel, then Miss May (Miriam's governess, who was later to teach Paul to paint) and Mr. Revell (the local non-conformist clergyman) gather round Mrs. Morel's bedside to discuss the new baby. The chapter ends at the top of page 88, and a new one begins immediately on page 88: Chapter IV, "Paul Morel's First Glimpses of Life", which takes him from the age of three to six. By page 90 he had reached four and Arthur was "three and a half" (corrected by Louie to "two and a half"), and Lawrence had begun to describe Paul's fits of motiveless weeping.

On 7 May (no. 263) Lawrence reported to Louie that he was on page 112 and found it all "very rummy"; then he reflected: "I don't think the last

chapter of Paul has action enough, moves sufficiently. It is the bane of my life, to get the action of a novel hurried along." The new chapter that he had begun on page 88 went on to page 125, so by the "last" chapter he must have meant chapter III, described above.

When Lawrence came to rewrite this chapter in Stage IIIa, his new chapter III contained a revised version of the same material. It began with the causes of Walter Morel's illness (now missing from Stage II) and continued with Jerry's visit to his bedside, his convalescence and the baby Paul's rejection of him, all modified but not changed a great deal. But the description of the birth of Arthur which followed was not only much abbreviated, but Lawrence even then deleted several lines of the new version at the end of the chapter (see Fragment 5). Evidently as he copied and modified the offending chapter, he identified its lack of movement with the scene at Mrs. Morel's bedside after the birth of Arthur.

At this point the paper evidence provides information of a different kind. Every letter that Lawrence wrote to Louie from 24 April to 14 May 1911 (nos. 256-268), excluding one (no. 262) written from school, was on A paper— with the odd exception of one that he wrote from Colworth Road and dated 29 April (no. 260). In almost all of them (except those of 11 and 14 May) he gave a progress report on the novel: clearly he sat down to both tasks in the evening in front of the one block of paper. The letter-paper and manuscript dovetail so neatly here that the one odd letter arouses suspicion. On closer inspection it is evident that Lawrence wrote it not on 29 April as he dated it but on 29 May. (He had started writing April on 26 May, too, but corrected himself in time.) The dates of the Whitsun holiday that he refers to, saying: "This time next week we shall be going to Eastwood", were 5-11 June. With this letter, Lawrence sent Louie his 'Paul Morel' manuscript: "It's nine o'clock. I'm going to do a bit of Paul. I send you this mass. I'm afraid it's heterogeneous; since I have never read it through, very blemishy. Correct it and collect it will you, and tell me what you think. This is a quarter of the book . . . What will you think of it, I wonder. I want to take it to Ada at Whitsun." This letter needs to be re-dated, then, 29 May, by which time Lawrence might well have written a quarter of his novel. The next day, 30 May (no. 272) he wrote Louie a postcard in French enquiring whether the parcel had arrived and apologising for not putting enough stamps on.

In what therefore turns out to have been the previous letter, of 26 May (no. 271), Lawrence had compensated for refusing Louie's request to see his manuscript poems, by promising: "But at Whit I will show you the first two hundred pages of 'Paul', that book of books." Then, having promised to show it to Louie personally, he changed his mind in favour of sending it for her to tidy up so that he could show it to his sister, Ada. He doubtless wanted helpful comments from both, but from neither could he expect the severe literary standards and independent knowledge and judgement of the material that Jessie Chambers had brought to bear on The White Peacock.[13]

Because Lawrence was letting Louie read his manuscript nearly a week earlier than he had originally promised, the "mass" he sent her may not have been the 200-page length he had predicted on 26 May; and since page 166 is now missing, it is tempting to speculate that it was lost because it was the last page in the parcel. Page 165 is the last of the A pages, and page 167 is the first of B paper, a thin 27-lined paper very like one he used for his letter to Louie of 1 June. The supporting speculation thus suggests itself, that Lawrence sent her all he had written when he finished his perforated pad of A paper. On neither page 165 nor 167 is there a natural break in the text: the one ends mid-sentence, the other opens mid-conversation; but at the end of this chapter, on page 176, Lawrence started a new one on the same page, so that did not provide a natural break either.

The two papers, A and B, are very different sizes, and while it is true that this alone could have caused the loss of page 166 at any time, it also suggests that the changeover may have seemed to Lawrence as suitable a break as any. Indeed, his evident haste, the suddenness of his decision to send the manuscript, his sense of its miscellaneity and his failure to put enough stamps on the parcel, all betray a degree of casualness that might easily result in loss of some sort.

Lawrence described this portion as a quarter of the book, and if it was pages 1-166 the expected total would be 664 small pages and nineteen short chapters. But since he failed to complete this draft, and the end of the next is now lost, and the completed final draft was found to be too long, it is impossible to guess what length he envisaged at this stage.

Chapter IV, "Paul Morel's First Glimpses of Life" (pp. 88-125), contained scenes from Paul's early childhood. Some were still present in the final draft, such as the evenings the children spent helping Walter Morel to make straw fuses while he told them tales of the pit; some were later dropped, such as a graphic account of Paul's participation in the games and candle-lit story-telling sessions of his elder sister Annie and her friends, and another of a day he spent on the ice in winter which led to an attack of pneumonia.

Chapter V, "Acquaintances" (pp. 125-176), described Paul's convalescence and his childhood friendships with "Ginger" and Miriam. In this version Lawrence placed Miriam in the household he later used for The Lost Girl, with an invalid mother, an impractical entrepreneurial father, and a governess, Miss May, who taught her to play the piano. The break caused by the loss of page 166 occurs just as Paul and Miriam are playing with Paul's rabbit, Adolphus, in her house. The chapter ended with elegant tea at Miriam's house followed by angry tales of his father's drunkenness on Paul's return home.

Chapter V (bis), "Launched" (pp. 176-207), was devoted to Paul's interview at Jordan's factory and his experience of working there. Again when Lawrence finished this chapter he began the next on the same page.

Louie did correct the manuscript for him a little, though there were many slips that she overlooked. In the whole of the surviving manuscript there are only fifteen alterations and corrections in hands other than Lawrence's, and

only eleven of them are unmistakably by Louie: they occur between pages 83 and 188. Her hand is distinctively even, rounded and upright or backward-leaning, and her alterations are minor one-word corrections of inconsistencies in names or dates, and small slips. She changed George to William three times (pp. 97,121); Job Arthur to Alec three times (p. 126); Muriel to Miriam once (p. 167) and three to two (p. 90). She also inserted till on p. 105, which was demanded by the context; wrote ois over Lawrence's spelling "chammy leather" (p. 188); and altered his "This was a lady" to "She was . . ." (p. 83). She used a thick black pencil from pages 83 to 126, and a very fine one from pages 167 to 188. If therefore, Lawrence had sent her pages 1-166, perhaps she had only reached page 126 by Whitsun, or had not noticed any errors thereafter, and then when they met during the holiday Lawrence gave her the continuation to read, which she marked using a different pencil.

The few changes made in other hands are hard to attribute.[14] Lawrence may have asked his sister Ada to correct any errors she noticed. Jessie Chambers read the manuscript months later, and some further pencil marks appear to be by her (see below).

After the Whitsun break, Lawrence gave up sending Louie regular progress reports on 'Paul Morel', but on 12 June, replying to Martin Secker who had written offering to publish his short stories, he humorously described his work on the novel (no. 274); "I stick at my third book like a broody hen at her eggs, lest my chickens hatch in a winter of public forgetfulness." But at some point during the summer he abandoned his post, probably during July (cf. nos. 280, 288) and possibly somewhere around page 353.

What he had written was very different in many respects, particularly in its plot, from the eventual novel, Sons and Lovers. Around the central core of the Morels' married life as their children grew up, he had constructed a variety of fictional narratives. He had introduced a girl, Miriam, whose teenage friendship with Paul caused him conflict with his mother, but had placed her in a petty-bourgeois family setting; he had divided his own character between two Morel sons, the artisan but artistic Paul who worked at Jordan's factory, and the clever but spoilt and volatile Arthur, a successful college boy. Then, using a shocking incident that apparently occurred[15] in his uncle's family, Lawrence described the manslaughter of Arthur by his father. Walter Morel died soon after his release from prison, and by the end of the eighth chapter, Lawrence had engineered the amicable isolation together of Paul and his mother.

In the next chapter he immediately began to introduce a thirty-year-old married woman, "a tall, silent blonde" called Frances who had left her husband, George Radford, after seven years of unhappy marriage. Some of the

later pages of this chapter describing Paul's conversations with Frances, Lawrence transferred to the next drafts as he rewrote the novel, changing the name Frances to Clara.[16] Thus the last pages of this draft that have survived are pages 341-351 now in the Final manuscript, and p. 353 standing isolated at the end of The Paul Morel Manuscript, its text filling the page with Paul and Clara's unfinished conversation. Although Lawrence did not complete this draft, page 353 itself contains no indication of where he left off.

On 17 July 1911 he had confessed to Louie that he had "not done any Paul lately" (no. 288), and in October he told her he had not done any for months (no. 316). The last twenty-five surviving pages are on D paper, which Lawrence also used for sixteen letters between 7 July and 20 October 1911. This might suggest that he renewed his efforts at the novel during the period from July to October, but it is unlikely, because for a long time he had been receiving attractive requests for short stories and poems, with the result that he was almost wholly occupied with stories from July to September and verse from early October. (This may explain why some of his verses for Love Poems are on D paper.)

Jessie Chambers and Stage II

During October Lawrence's publisher, Heinemann, began pressing him to promise that 'Paul Morel', which Lawrence said was "half done", would be finished "for March" (no. 323). But when he turned to it again, he must have despaired of breaking through the block he had developed towards it during the summer, for instead of completing it, he suddenly decided to send the unfinished manuscript to Jessie Chambers and ask her opinion of it (E.T., 190).

Relations between Lawrence and Jessie were very troubled during this period, and indeed were effectively at an end six months later. They had barely been in touch since they had met at the time of Mrs Lawrence's funeral in October 1910 and quarrelled about Lawrence's recent engagement to their mutual friend, Louie (E.T., 183-184). Lawrence had sent Jessie a complimentary copy of The White Peacock on its publication in January 1911, in gratitude for her help in writing it (E.T., 189), but he left her to discover from another mutual friend, Helen Corke, that he had started an autobiographical novel (E.T., 190) which was bound to contain reflections on his relationship with herself. It was through Helen Corke that Jessie met Lawrence again at this time: for it cannot have been by chance that he and Helen Corke both took their guests—his brother George, and Jessie, respectively—to the same performance at the Savoy Theatre in London on Saturday 7 October 1911.

What Lawrence then sent Jessie, therefore, "some time" after this meeting, (E.T., 186), was The Paul Morel Manuscript now at H.R.C., together

with the opening and latter pages which are now missing. Jessie had played a crucial rôle in Lawrence's first publications, both by sending his poems to the English Review (E.T., 156-159) and by "nursing" his first novel, The White Peacock, into existence (she records with some pride his tribute: "I its creator, you its nurse.") (E.T., 189). But in writing The Paul Morel Manuscript Lawrence had sought not only to exclude her but to replace her in her rôle of confidante and adviser with his fiancée, Louie Burrows, his engagement to whom Jessie had disapproved: "I told him he had done wrong to involve [Louie] in the impasse of our relationship" (E.T., 183-184). Therefore, even if Jessie felt it a pleasure or at least a welcome duty to respond to Lawrence's appeal for comments on his novel, she must to some extent have been predisposed against the manuscript.

Twenty years later she included a critique of it in her memoir, E.T., which would be the sole account of this draft in existence had The Paul Morel Manuscript not been preserved; but now her description can be read in conjunction with the original document itself.

In a strongly written paragraph she summed up its inadequacies: its deadlocked, "story-bookish" plot (a word she had used of the first draft of The White Peacock), and its "tired writing" which convinced her that Lawrence "had had to force himself to do it" (E.T., 190-191; cf. 116). Her phrases have an uncanny resonance with the remarks quoted above that Lawrence had made in his letters to Louie as he wrote this manuscript. And yet her judgements are not so pervasively true as her emphases would have them: much of the surviving manuscript displays Lawrence's characteristic vitality.

More importantly, it is now evident that her account then continues with some curiously inaccurate statements: for, passing over in silence the father's manslaughter of Arthur in order to home in on Lawrence's treatment of "the mother's opposition to Paul's love for Miriam", Jessie wrote (E.T., 191):

Not one of these is an exact quotation from The Paul Morel Manuscript. This cannot be explained by the possibility that Jessie was quoting from parts of the manuscript now lost, for the central chapters devoted to Paul's relationship with Miriam have survived. The sentences that correspond most closely with Jessie's "quotations" are:

How did this baffling discrepancy arise? There are three possible explanations: either she wrote this entirely from memory, or she used very brief notes made at the time, or she was inadvertently quoting from notes she later made of the Stage IIIa manuscript, the whole of which she read. The latter two alternatives are rendered doubtful by Jessie's declaration in a draft of her memoir: (Delavenay, 710) "I had to make a new start in life . . . So I found it was necessary to break every link that had bound us together, and I destroyed as far as I possibly could every line of his writing. I burned all the letters that I had kept, and even tore the fly leaf out of several books where he had copied poems for me." However, she may have failed to destroy everything or have saved, for example, a diary containing reflections or notes which she was able to turn to when writing her memoir.

Too little of Stage IIIa survives to rule out the possibility that her quotations accurately represent sentences Lawrence had written there, but in the few remaining passages with which her "quotations" do have some correlation, the wording is even less close than that of the passages in Stage II. They are[19]:

and:

When the equivalent passages in the published form of the novel are looked at,[20] it is clear that Lawrence developed the Stage II versions of them

with a certain consistency in each revision: he steadfastly moved away from the suggestion that Miriam could win Paul's sex sympathy. This makes the third explanation of the discrepancy the most implausible. If, on the other hand, Jessie's "quotations" from Stage II represent the way in which she remembered at the age of forty-four something she had read at twenty-four, the gist of the sentences that had struck her had been transmuted in her memory into explosive and impatient formulations with the passage of time. In view of the fact that, as I explain below, the manuscript was only in her hands for two weeks, or at most just under a month, it seems to me marginally less likely that she would have developed sufficient familiarity with these passages to carry them in her memory for twenty years, and rather more likely that she had kept some sketchy notes. That would also give her the confidence to offer her sentences as quotations, since they would then be at least anchored in the key words which she had jotted down at the time.[21]

Jessie did not write comments in the margins of The Paul Morel Manuscript as she did later in some parts of the next draft, perhaps because she knew Lawrence intended to rewrite it—or because she was not driven by it to the degree of anger and distress that was provoked in her by the next draft. But it must have been she who pencilled (non-punctuational) brackets around Lawrence's description of Miriam as a "woman who exists only to plunge herself into the mystical dark of life" (p. 246)[22] perhaps as a description she intended to challenge. In her memoir, however, she indicates that her reply consisted of more general comment and advice: (E.T., 191-192):

Lawrence may not have been expecting such radical advice, but, as Jessie explained to the French scholar, Emile Delavenay (670): "He took me at my word, and entirely recast his novel. He included also the story of our household, which I had not suggested." She did not specifically ask Lawrence to set the figure of Miriam in a portrait of the Chambers's home, but that was the

effect of her insistence "true to life". Her concern in pressing for that seems to have been a desire for Lawrence to see, as she did, that the death of Ernest would prove a portent for Bert, too, should he prove equally unable to overcome his "strange obsession" with Mrs Lawrence (E.T., 192).

I am pressing her remarks to conclusions which she deliberately—or delicately—resisted in her memoir (E.T., 192-193). She wrote of The Paul Morel Manuscript: "The elder brother Ernest, whose short career had always seemed to me most moving and dramatic, was not there at all. I was amazed to find there was no mention of him" (E.T., 191). She had forgotten the William of this draft, who appears to represent Ernest as a child, and whose name Louie had occasionally had to correct from "George" (the actual name of Lawrence's other brother). This oversight is not as strange as her failure to mention the Arthur of this draft. The astonishing climax of the plot, in which the son's vituperations of his father whip Walter Morel into such a blind rage that he hurls a poker at his son, and kills him, would be the first thing that most readers of this manuscript would mention in a summary. Jessie must have known the true basis for the story; presumably she found Lawrence's use of it melodramatic, or was not interested in his exploiting alternative narrative ploys for developing his subject. Her emphatic "no mention" of Ernest was related to her emphasis on his "short career": by comparison with the death of Ernest, which rendered his life-story "moving and dramatic", "poignant and interesting" and "vital", she must have found the death of Arthur "story-bookish" (the sentences about Ernest quoted above follow immediately upon the use of that word).

Jessie's silent amazement at Lawrence's fiction of the death of a "brother", and the focus it gave to her advice, produced a radical and thought-provoking change in the novel. In the revision, the father is no longer the agent of his son's death, and the hatred for the father shown in Stage II thereby loses its central vehicle; the college boy is no longer represented, nor are the bitter denunciations with which he had whip-lashed his father.

Stage IIIa: the start

In Letters I Professor Boulton has dated Lawrence's letter to Jessie accompanying his unfinished manuscript of Stage II (no. 324) after the second of his two interviews at Heinemann's, on 5 and 20 October 1911, which appears from Lawrence's account (no. 323) to have been the more businesslike and specific. The fact that Lawrence and Jessie met on Saturday 7 October might suggest an earlier date; but Jessie herself could not remember whether the parcel arrived in October or November (Delavenay 670). The later date gives her just a fortnight to return the novel with her comments before 3 November 1911, when Lawrence started Stage III. For it is most unlikely that he began the new draft until the old had been returned, not only because he was waiting for Jessie's suggestions before deciding how to proceed, but also because, as is evident from the comparisons that can be made, Lawrence's method of rewriting was to place the discarded draft in front of him and transcribe it with modifications.

On 3 November he wrote to Louie with the air of one putting on a yoke (no. 330):

This letter, like the two previous ones to Louie, was written on e paper, and when Lawrence had finished writing it he started his novel again as he promised, writing the first words on the same pad of poor-quality lined paper. For there survives an abandoned opening of the novel on seven unnumbered pages of this paper, which is now kept together with The Paul Morel Manuscript at the H.R.C. Being evidently the novel's opening, these pages are placed at the front of 'Paul Morel' Stage II, but in fact they post-date it by several months.

The text runs out nearly half-way down the seventh page, a point at which Gertrude Morel had finally learnt that the young man, "a school teacher who was a good Latin scholar," whom she had loved at twenty, had subsequently "married his landlady, a woman of forty odd years, who had money". Lawrence had arrived at this point in a brisk 150 lines from the opening: "'The Breach' took the place of Hell Row. It was a natural succession. Hell Row was a block of some half-dozen thatched, collapsing cottages . . ." It is clear from the textual development that these seven pages preceded the two other versions of the opening which also survive: Stage III proper, partly in the Fragments and partly in the Final manuscript, and Stage IV in the Final manuscript. The shape remained the same, but by the time Lawrence had finished working it up, he took 310 lines to reach the same point.

Lawrence was ill with pneumonia from 19 November 1911 till the end of the year, and convalescent until February. As mentioned above in the section on Paper use in Lawrence's letters, the evidence from his letter-paper is unusually clear-cut in this period.

In summary, he used f paper, a fine quality paper, watermarked "Court Royal" with a shamrock, rose and thistle design, for his letters from 15 November to 13 December 1911. (He was so ill that some of these letters had to be written for him.) Then he moved on to another good quality paper, "Classic Parchment", which he used for all his letters from 17 December to 7 January, and finished later in Eastwood (after his return from convalescing at Compton House, Bournemouth), using it for every letter from 10 to 28 February and then intermittently until 17 April 1912. No "Court Royal" (f paper) occurs again after 13 December 1911.

This paper evidence leads to a number of conclusions:

1. The fact that Lawrence abandoned his first attempt of a draft of Stage III at the seventh (unnumbered) page, and began again, was not caused by his illness.

2. He began again on f paper, copying and modifying this seven-page

draft, and proceeded to write pages 1-74 of Stage IIIa on f paper between dates after 3 November and before 19 November, when he fell ill. (On 15 November he wrote to Louie—on f paper—that he was "frightfully busy", "rushed", "squashed with work" and exclaimed: "Poor Paul, I don't know when he will be done . . . Now I will to Paul." no. 338)

3. At this point, because Lawrence was too ill to continue with this draft of his novel, his supply of f paper was assigned for letter-writing, and it was used up during the next month.

4. At page 74 (see Fragment 5) Lawrence had reached his satirical description of Jerry at Walter Morel's bedside about to produce the bottle of beer: page 74 ends mid-way through the paragraph:

If not at the end of it, Lawrence must have been somewhere on page 74 of Stage IIIa when he fell ill.

5. The first surviving page of the previous draft, Stage II, page 72 of The Paul Morel Manuscript, has the same one-sentence paragraph on its fifteenth line: Lawrence was clearly copying and modifying The Paul Morel Manuscript to produce the new draft.

6. This explains why The Paul Morel Manuscript survives from page 72 onwards, or at least why pages 1-71 are lost: this is the point at which Lawrence had stopped when he fell ill in Croydon. The earlier pages of Stage II, which he had already copied, he no longer needed and so presumably he left them behind or threw them away when he left Croydon for Bournemouth. The rest of the old draft he took with him back home to Eastwood so that he could continue copying it with modifications from the point at which he had left off.

7. This suggests that the lost seventy-one pages of the opening of Stage II were similar to the opening of Stage III, now reconstructable from the Fragments and Final manuscript. Certainly the two versions of the episode between Walter Morel and Jerry, which can be compared, are very similar.

8. One theory which suggests itself concerning the renumbering of pages 72, 73 and 74 of The Paul Morel Manuscript must be rejected. Because of the way Lawrence later transferred pages from an earlier draft to the current one instead of copying them out again, it might be thought that the altered numbers of these first-surviving pages of Stage II indicate that Lawrence had been tempted, on taking up Stage III again at pages 74-75, simply to incorporate the old material rather than write it out again, but that after three pages he changed his mind. This cannot be the case, partly because he did not delete the first fourteen lines of page 72, as he would have needed to do,[23] but mainly because these pages must have been first 74, 75 and 76,

and renumbered 72, 73 and 74 (the last of them is particularly clear, as the original 6 is revised into a very a-typical 4) which is the wrong way round, even supposing Lawrence had made the error of altering the first of the three to 74 instead of 75, and so on.

9. Finally, it follows that when Lawrence took his novel up again between 9 and 23 February 1912, and said in a letter of 23 February (no. 392): "'Paul Morel' is going pretty well, now that I have once more tackled it," he had not begun from the beginning again, but took it up at page 75 on g paper—which, alas, does not appear in his letters, by way of tidy corroboration, because he was now using up his supply of "Classic Parchment".

Stage IIIa: the completion

Back in November, when Lawrence had first received Jessie's suggestion that he rewrite the novel keeping it "true to life", he had responded by asking her to write down some reminiscences for him to use (E.T., 193). He did not need these for revising the opening seventy-four pages in Croydon, but soon after his arrival in Eastwood on 9 February, and probably before 23 February when he reported himself underway with the novel again—"early in the week" as Jessie recalled (E.T., 195), and therefore possibly Monday 12 or Tuesday 13 February—he visited her to collect her notes.

On 4 February Lawrence had terminated his engagement to Louie, and he now wrote Stage IIIa with Jessie's moral support as he had written Stage II with Louie's. But the writing of this draft was more rapid, and as he lived nearer to Jessie he was able to give her pages of manuscript to read continuously as he wrote it, delivering them first by hand and later posting them.

On 23 February Lawrence officially resigned from his teaching post, and he now worked single-mindedly on this novel, turning away nearly all other offers of publication, whether short stories or poems. There were, however, a few other things that he allowed himself to do. From 2-8 March he took a trip to visit some old friends including the Daxes in Shirebrook.[24] From 25-31 March he was in Bradnop, Staffordshire, in fulfilment of a promise to visit his old school-friend George Henry Neville. (During this period he also met, or re-met, Frieda Weekley.)

All the while he continued to work hard at his novel, even taking it with him when he went to Shirebrook: there are letters on j paper on 1 March from Eastwood (no. 396) and 4 March from Shirebrook (no. 397), and on k paper on 6 March from Shirebrook (no. 399) and on 8 March from Eastwood again (no. 400), so presumably he wrote those parts of his novel that made the transition from the one paper to the other in Shirebrook.

His sense of urgency derived firstly from his determination to fulfil as nearly as possible his deadline for delivery to Heinemann, which he now set at May. Secondly, he had received in January while ill an invitation to visit

relations in Germany, which had apparently given him the idea of writing a book of travel sketches,[25] and he had at once set his heart on going in the spring. When it was that he also began to think of this trip as an opportunity for an affair with Frieda it is not possible now to know for certain.

There are not the same detailed accounts in Lawrence's letters of his progress with this draft that there had been for Stage II, partly because of his direct contact with Jessie Chambers, and partly because she later destroyed his letters. Instead, there is an unusually full, and, in a sense, first-hand reminiscence in Jessie's memoir, written from her special perspective.

At first sight there seems to be a discrepancy between her account and the impression given by Lawrence's letters concerning the date at which he completed this draft. Jessie recorded that it was written "in about six weeks" "at a white heat of concentration" (E.T., 204, 201) and implies it was finished before 25 March 1912. Lawrence's letters indicate that it was finished about 11 April, which was seven weeks from 23 February (the latest date by which he had taken it up again).

In describing how Lawrence brought pages of the novel to her as he wrote it, Jessie does not state that he collected the pages she had already read, but gives the impression instead that he left it all with her to accumulate—as he must have done when he stopped bringing it personally and began to send it by post. This latter change may have had something to do with his trip to Shirebrook, but Jessie was convinced his motive was solely embarrassment at what he had done with her reminiscences, and, in effect, what he had written about herself (E.T., 201). When she had read the last pages, she proposed a date for a meeting, but Lawrence replied that he was already committed to visit Neville: since that visit took place from 25-31 March, he had, according to Jessie, completed the manuscript sometime before 25 March (E.T., 205). This is rather surprising in view of the fact that only on 13 March Lawrence had been predicting to Heinemann's editor, Walter de la Mare, that it would take him another month.[26] The most satisfactory interpretation is that Lawrence

was allowing himself time for an immediate read-through after he had written to the end, and after Jessie had, like Louie, "corrected and collected" the manuscript and told him her opinion of it.

On Monday 1 April, the day after his return from Bradnop, Lawrence visited Jessie to collect the novel and hear her comments. But she was so overwhelmed by the injustice she felt he had done her by his portrait of Miriam, that she declined the inevitable quarrel, and, instead, "put some notes in with the [parcel of] manuscript" (E.T., 210-211). That parcel must have contained the twenty-three pages that now survive annotated with Jessie's objections and revisions, together with four pages of commentary entitled Chapter IX in which she expounded her criticisms, and at least one of the episodes written out by her.[27]

Thus the apparent discrepancy between E.T. and Lawrence's letters may be explained: when on 3 April he wrote to Edward Garnett (no. 411), "I shall finish my Colliery novel this week—the first draft. It'll want a bit of revising", he was presumably "finishing" it in the sense of pondering Jessie's criticisms and perhaps rewriting chapter IX—now too disfigured to submit it to a publisher—and tidying up the other pages she had annotated.

For example, on page 311, which was later transferred to the Final manuscript (where it became page 352), Jessie had pencilled in a correction of fact on the bottom line, overflowing to the top line of the next page. Lawrence did not rewrite the pages but penned his slightly different revision over hers so as to obliterate hers. His text had first read:

Jessie's correction, which is hard to read now, was intended to replace "a thousand times": Lawrence overwrote this in ink changing the first two sentences to:

On 11 April, after a week of (I suggest) such work on his manuscript, he reported to Walter de la Mare at Heinemann's (no. 415): "I have finished in its first form the colliery novel. Now I want to leave it for a month, when I shall go over it again. There are parts I want to change. Shall I send it to you for your opinion now at once, before I do any revising, or shall I pull it closer together before you see it?" There is nothing in the surviving

'Miriam Papers' to suggest that Jessie had advised him to pull it together: perhaps this was the result of his reading it through during this week. In the event, he took it to Germany on 3 May and revised it there.

Lawrence called Stage IIIa the "first draft" and the "first form" of the novel because he had recast it so radically it no longer resembled Stage II, and because now that he had produced the first version of the whole in the form it was ultimately to take, he wanted to work it up. Unfortunately the end of this version is now lost, but Lawrence did give a clue as to how long he expected this draft to be when he wrote to Edward Garnett from Shirebrook on 6 March 1912 (no. 399) that he had "done two thirds or more". This letter was written on k paper, and the changeover from j to k paper, which happened in Shirebrook, took place between pages 281 and 309. If this was two thirds, the whole would be about 450 pages. (By contrast, the Final manuscript was 533 pages, much of it written much smaller on much larger paper.)

The remains of Stage IIIa

Very little of Stage IIIa now survives: 139 pages in seven sequences on six different batches of paper (not including the sections of D paper transferred towards the end from Stage II). These are pages 1-85, 182-192, 204-226, 272-281, 309-315, 317-318, 328-331 (the D paper of Stage II became pages 332-333 and 337-347). The papers are f to l, of which f, j and k only are found in the letters.

The largest sequence is the opening eighty-five pages on f and g paper, constituting chapters I-III and the opening of chapter IV, of which the first seventy-four pages are the "Court Royal" pages which Lawrence wrote before his illness. The opening chapters are the part of Stage IIIa that Lawrence seems to have copied most closely from Stage II, but because the first 71 pages of Stage II are now lost, and pages 86-181 of Stage III are also lost, there are now only twenty pages of close overlap between these two versions: Stage II pp. 72-91 and Stage III pp. 74-85. However, the fact that certain episodes from the beginning of Stage II are also found in Stage IV—notably the making of straw fuses and Paul's interview at Jordan's—means it is very probable that despite the eventual divergence of plot, the overlap between Stages II and III continued beyond what now survives of Stage III.

This is particularly significant because of another curious inaccuracy in Jessie Chambers's memoir. In her description of the beginning of this draft, she selects for praise, as if now written into the novel for the first time, two passages that she must have read (very little different) in The Paul Morel Manuscript (E.T., 197-198):

However much one may agree with Jessie's praise here—leaving aside her failure to notice Lawrence's ironies at the expense of Jerry and Walter Morel (which included that the latter's illness was caused by drink and not by an accident at the pit)—she implies that it was in this draft, written with her advice and moral support that Lawrence first described those everyday things she had not noticed before. But her examples were both present in The Paul Morel Manuscript, and the description of Mrs Morel ironing (which survives intact amidst radical alterations in the Final manuscript)[28] was very probably, like the other scene, copied with modest revisions into Stage IIIa. In view of the fact that Jessie wrote equally forcefully of the stale, tired writing of Stage II, her emphatic praise here must reinforce the suggestion made above that she had been predisposed against that manuscript.

Chapter I, "Antecedents", has survived complete in its Stage IIIa form: pages 1-43 (of which twenty-four pages were eventually retained in Stage IV). It was similar in outline to the revised chapter I in Stage IV (that is, largely as published), but Lawrence had not yet introduced the small boy William's arrival for lunch and visit to the Fair, or Walter Morel's return home after helping at the public house and his brief conversation with his wife before the authorial voice resumes the story of her antecedents. In Stage IIIa, while Mrs Morel waited for her husband to return from his holiday trip to Nottingham, Lawrence offered some authorial reflections on her disappointment with her marriage, such as:[29]

followed by descriptions of the stifling kitchen and the relative emotional states of the two partners that evening before they met and quarrelled, such as (Fragment 2, p. 34): All of this Lawrence dropped at Stage IV in favour of the dramatic inspiration of having Morel arrive just as his tired and heavily pregnant wife is engaged in the effort of pouring hot liquid from a large saucepan.

Chapter II, called "Birth and Death", then "Birth and Horror" and then simply "Birth", can also be reconstructed almost intact at Stage IIIa: pages 44, 46-67. This was the only one of the first three chapters that Lawrence modified at Stage IIIb. Eight pages from Stage III (a and b) were eventually retained in Stage IV, but the rest of this version of the chapter was only revised a little as it was copied into the Final manuscript. For example, at Stage III Lawrence had not yet had the idea of describing Morel working at the pit before coming home to discover that Paul had been born; he also introduced in Stage IV some remarks about the older boy William's distress at his mother's pain and his dislike of the new baby—but these were deleted by Garnett when he pruned the Final manuscript. Finally, at Stage IIIa chapter II had ended with Walter Morel's escape from his family to the local pub on the Sunday evening following the Saturday night row in which he had struck his wife [with a drawer]. (It was at Stage IV that Lawrence added the semi-comical coda of Morel's half-hearted attempt to leave home with a bundle of possessions.)

Chapter III, numbered at Stage IIIa pages 68-82 and called first "Aftermath" and then "Morel Reaps the Whirlwind", was all rejected and rewritten at Stage IV, and is now entirely found among the Fragments. As described above, it opens with the causes of Walter Morel's illness, and continues with a revision of the Stage II version of chapter III: Jerry's visit, the baby Paul's rôle in disturbing the brief harmony between Mr and Mrs Morel during Morel's recuperation (which was re-used in Stage IV but deleted by Garnett) and finally a severely abbreviated account of the birth of Arthur. (A very brief remark at the end of the chapter, on the young life of William, was the seed of expansion to many pages in the Final manuscript—of which Garnett deleted a great deal.)

Chapter IV, "Glimpses of Early Life", of which only three pages now survive (Fragment 6 pages 83-85) is also a revision of Stage II. It begins with the successful relationship between the handsome baby Arthur and his father, contrasted with the sickly and colourless Paul who followed chattering at his mother's heels all day; and the pages run out at the familiar account of Paul's fits of motiveless weeping. All this material occurs in modified form in chapter III of the Final manuscript, where the chapters are fewer and longer.

As one looks from the opening three and a bit chapters of Stage IIIa back to the earlier versions—the overlapping first twenty extant pages of The Paul Morel Manuscript (pp. 72-91) and the seven unnumbered pages of the abandoned opening—and then forward to the much revised final version of the work, certain general impressions emerge. Lawrence's first inspiration seems to have been the single thread of his mother's jarred existence in the alien, recently built mining village, with her memories of her betrayed youthful and romantic optimism still very vivid to her; and he only gradually diversified that rather static narration of bitterness with the dramatic accretions of the lives of other members of the family, particularly William. Jessie's recommendation with regard to William bore fruit progressively: aspects of his youth had already been included in Stage II; his death was now entered at her suggestion into Stage IIIa (see the account of h paper below); but further aspects of his childhood and early manhood were introduced as Lawrence rewrote the novel. It seems that Lawrence's perception of William's rôle developed along the lines Jessie had had in mind slowly,[30] and accumulated momentum as he rewrote the novel; and it might be added that Garnett made an error of perception when he pruned much of that growth in the early pages.

In Stage III Lawrence's tone towards Walter Morel was more lighthearted, Mrs Morel's resentment of her husband was occasionally alleviated by humour and Lawrence's treatment of both was more even-handed at the outset of the narrative. Then, although the movement of the later drafts was less static, had more of the action that Lawrence found it so hard to instil, the bitterness became more relentless, the humour more scornful.

From this point on, the surviving pieces are more fragmentary, and it is here that the identification by paper-types is most useful.

Of the two sections of h paper, the first, pages 182-189, narrates the death of William from the moment of Mrs Morel's arrival at his bedside, as far as Paul's subsequent illness. The fact that these pages, written for the first time now in response to Jessie's advice, were eventually incorporated into the Final manuscript shows that Lawrence established them largely to his satisfaction at the first writing. (The narration was completed on three pages of g paper, 190-192; see Table 2.)

The second sequence on h paper is the twenty-three page section of chapter IX, pages 204-226, the whole of the chapter except its opening pages. This is the longest surviving sequence of Stage IIIa apart from the first three chapters, and is the equivalent of chapter VII of the published novel: it relates Paul's teenage friendship with Miriam and was rewritten because of Jessie's many objections.

The part of Stage IIIa on j paper, pages 272-281, which constituted the end of a chapter and survives because Lawrence found it satisfactory enough to transfer bodily into Stage IV to be the end of chapter VIII, recounts the terrible and pivotal scene of Paul's quarrel and reconciliation with his mother:

It was not the first time Lawrence had written up this exchange, for he was here further developing it from The Paul Morel Manuscript and A Collier's Friday Night (Act III). In other words, this scene was one of the foundation episodes present in all the versions of the novel for which we have any evidence, and must have been central to Lawrence's conception.

The two sequences of k paper were numbered 309-315 and 317-318 at this stage. But when Lawrence re-used them in the Final manuscript he did so in reverse order, renumbering them 350-355 and 345-346. The first, at Stage IIIa, contained Paul's twenty-first-birthday letter to Miriam in which he had called her a nun, and his reply to her reply. It was his description of Miriam's reaction to the first letter that Jessie had corrected on page 311. The other, two-page, passage, which at this stage followed Paul's letters, was the latter part of a scene of partly dare-devil and partly passionate flirtation between "Beat" and Arthur Morel. Evidently Lawrence decided at Stage IV to make the virile portrait of Arthur in this scene (followed by some sketching of the changes at home after Annie's marriage and a description of the physical contrast between Clara playing at the farm and Miriam standing aloof) lead up to, and so set a pointedly contrasting tone before, Paul's precocious-juvenile, intellectual-spiritual letters to Miriam. With the second of them he then brought chapter IX of the Final manuscript to a close, as "the end of the first phase of Paul's love affair".[31]

In the four pages of l paper that follow shortly (later transferred into chapter X of the Final manuscript) Paul invites Clara to apply for an overseer's job at Jordan's, and this leads straight into a section of pages taken from the Stage II draft, The Paul Morel Manuscript, in which the growing friendship between Paul and Clara (her name now changed from Mrs Frances Radford) is quickened by their walk to the Castle on his birthday and her subsequent gift. These had been pages 337-351 of Stage II and they were renumbered when transferred to Stage IIIa, pages 332-347. Thus our knowledge of Stage IIIa ends precisely where our knowledge of Stage II ended.

The remains of Stage IIIb

Only twenty-two pages first introduced into Stage IIIb now survive: pages 45, 46a-49a, 62, 68-69 on n paper which Lawrence used for his letters between 29 June and 18 July 1912, and pages 353-355, 358-362, 370-375 on m paper which he used for letters on 16, 17 and 21 May 1912.

This was the period of Lawrence's sojourn in Germany. He was abroad for the first time in his life; his affair with Frieda had produced tumult with her husband and her family and turmoil within Lawrence himself. After a week abroad, he went to Waldbröl to stay with his Krenkow relatives,[32] who had issued the original invitation to him to visit Germany. He was in Waldbröl

only from 11-24 May, and on 16 May he settled down to revise his novel.

His letters from here on 14 and [15] May (nos. 443,444) are on a 26-lined paper which does not appear in the novel, but from [16] to 21 May (nos. 445, 446, 447) are all on m paper—otherwise he wrote postcards, and probably one letter to Jessie. She reported in her memoir that he wrote from Germany saying (E.T., 216): "I am going through Paul Morel. I'm sorry it turned out as it has. You'll have to go on forgiving me."

On m paper he wrote fourteen new pages which he inserted into his draft and which survived into Stage IV as parts of chapter XI, "The Test on Miriam", some of them very little altered in Stage IV. It is impossible to tell whether these new pages contain quite new material or just extensive revision of Stage IIIa writing, but they are particularly interesting in that they convey some of the first of Lawrence's reflections on his relationship with Jessie under the impact of his very new relationship with Frieda.

Pages 353-355, renumbered 397-399 in Stage IV, contain Paul's conversation with Miriam when he went back to her, about their "marrying", from "'I can't marry you,' he continued slowly" to "As she stood under the drooping thorn-tree." Pages 358-362, renumbered 402-406 in Stage IV, contain the cherry-picking sequence followed by the walk in the pinewood, from "There was a great crop of cherries. . ." to "Instinctively, they all left him alone." Pages 370-375, renumbered 414-417 and 419-420 in Stage IV, contain Paul's final breaking-off with Miriam, starting "'Well,' said his mother, 'I think it will be best,'" and ending "His mother looked at him." But this section was altered a great deal at Stage IV, with a whole page inserted (p. 418) towards the end of the bitter dialogue between Paul and Miriam, after "never fight you off" and before "He hacked at the earth till she was fretted to death."

On 23 May he announced that he had "finished all but 10 pages" (no. 453); and the next day he left Waldbröl for Munich and his "honeymoon" with Frieda. From 1 June for two months they settled in Alfred Weber's apartment in Icking, near Munich, as Lawrence described to his Eastwood friend Mrs Hopkin on 2 June 1912 (no. 457): "Now Frieda and I are living alone in Professor Weber's flat. It is the top stor[e]y of this villa—quite small—four rooms beside kitchen. But there's a balcony, where we sit out, and have meals, and I write."

It may have been here that Lawrence tinkered with chapter II of his novel; for he inserted revised pages in chapter II on n paper, an extremely transparent cheap paper, rather like grease-proof paper, which he also used for Once written between 2 and 29 June, and for his letters of 29 June and 3, 8 and 18 July. As this paper is not found outside these dates, and as Lawrence used different paper for revisions to the end of the novel, it is likely that he made these changes to the second chapter nearer those dates. Moreover, he talked on 2 June (no. 458) of sending 'Paul Morel' off to Heinemann, but did not send it till 9 June.

On this paper he inserted five new pages between pages 44 and 46 of his previous draft (Fragment 3, pages 45, 46a-49a) in which he described Walter

Morel cleaning up for his wife during the latter part of her pregnancy, and the gossip between the women and the traders in the alley of the Breach. He also tried but did not complete a two-page revision to the end of chapter II (see Fragment 4, pages 68-69—this explains the reduplicated numbers which perplexed Professor Schorer). The only n page that lasted into Stage IV was page 62 (renumbered 64 in the Final manuscript), a one-page replacement among his IIIa "Court Royal" pages, on which he described Morel throwing a drawer at his wife.

As these different strata fall into place in the final morphology of the novel, it is not uncommon for material transferred from earlier stages to be found at the end of a chapter. This gives the impression that Lawrence knew the destination at which he intended each phase of his narrative to arrive but took time establishing the route. Yet where the different drafts can be compared intact, Lawrence can be observed copying out again almost verbatim whole episodes. For example, in the visit of Jerry to Walter Morel's sickbed and most of the ensuing twenty pages of overlap between Stages II and III, the revisions Lawrence introduced as he copied could have been entered on the original Stage II text without much disorder, and Lawrence could have transferred the pages, as he later did elsewhere. Thus it would be a mistake to suppose when Lawrence rewrote a passage afresh onto new paper that he was compelled to because of the quantity of revision, or that it was entirely new material. It may be that, particularly as he wrote Stage IV, his zeal for making fair copies flagged towards the end of a chapter. Nonetheless, it may be significant that chapter-ends often remained unchanged from Stage III to Stage IV, and that the surviving sequences of Stage III often had been or became chapter-ends.

Edward Garnett's notes on Stage IIIb

By contrast with the opening chapters, the latter part of the novel was not only not fixed early on, but expanded considerably in the final draft. Whether or not this was the effect of Garnett's detailed notes on the Stage IIIb version of the novel which Lawrence had before him as he wrote Stage IV, the resulting length of the Final manuscript was clearly not what Garnett had intended.

It took Heinemann only three weeks to reject Lawrence's novel outright at the sight of the Stage IIIb manuscript, and send it back (on 1 July 1912). His editor, Walter de la Mare, wrote almost at once to Edward Garnett, editor for Duckworth, to explain the situation (Letters I, 424 n. 1), with the result that Garnett asked Lawrence for the manuscript. Heinemann's main objections were that the work lacked "unity" and "reticence" (Letters I, 421, n. 4). Despite his anger at the rejection, Lawrence apparently accepted the former point, for on 8 July he told Garnett, who had given him similar help with The Trespasser (no. 467): "I will make what alterations you think advisable. It would be rather nice if you made a few notes again. I will squash the first part together—it is too long."

It took Edward Garnett only three weeks to accept Lawrence's novel in

principle, and to return the manuscript to him by 22 July together with notes which Lawrence described as "awfully nice and detailed." (no. 472) These notes have not survived, but by chance some of Garnett's jottings toward them have, tucked away among Lawrence's letters to him in the Berg Collection at the New York Public Library where they and their significance have hitherto been overlooked. They are written in terrible scrawl—Garnett's handwriting is not the most legible at the best of times—on the back of a piece of paper[33] which Lawrence had evidently slipped in the parcel with the manuscript when he sent it to Garnett on 4 July, bearing in large clear letters his address in Germany.[34]

This is one final piece of evidence about the composition of the novel which the identification of the remains of Stage III by paper analysis makes it possible to set in place. For three of the page numbers beside the sixteen brief notes on this piece of paper refer to surviving pages of Stage III. The first is: "331 'Fabian v. Socialistic'". This page became page 371 of chapter X, "Clara", in the Final manuscript, and Lawrence had written of Clara at Stage IIIa: "She had acquired a certain amount of education, under Fabian and Socialistic and Unitarian influence." In Stage IV he completely revised the middle part of this page, beginning with that sentence, and wrote instead: "During the ten years that she had belonged to the Women's movement, she had acquired a fair amount of education. . ." Lawrence's revision here confirms that although this scrap of jottings remained among Garnett's papers, he sent Lawrence notes based on them.

Another jotting refers to page 341: "Talk is a little cheap." This page was one of an eleven-page sequence carried forward from Stage II and renumbered 337-347 in Stage III; page 341 was then renumbered 381 in Stage IV. It was part of the description of the walk Paul and Clara took to the Castle in Nottingham on Paul's birthday: their actual "talk" lasted from page 341 to 345 (renumbered 381-385 in Stage IV). On page 341 in Stage III they merely commented on how small the people in the streets below the Castle seemed from the Castle grounds, and Lawrence appears not to have found anything "cheap" to rewrite, for he had not altered this or the next few pages very much. However, he did revise the bottom half of page 342; but when Garnett eventually saw Lawrence's revisions, in the Final manuscript, in December 1912, he still found the passage unacceptable, and deleted all the conversation from half-way down page 342 to the end of page 343 (pages 382-383 in the Final manuscript), so that it was never printed. In these 35 lines, Paul turned the conversation round to talk about Clara's apparent self-loathing, and then about her husband Baxter's "naturalness".

The other page that corresponds with Garnett's jottings, page 373 (renumbered 417 in Stage IV), contains the words "like a cynicism" which Garnett found "lurid" (indeed they are underlined with Garnett's wavy line in pencil in the manuscript). This page was one of the pieces of blue paper that

Lawrence used in Waldbröl for writing Stage IIIb. In that May 1912 version he had written:

In response to Garnett's criticism, Lawrence heavily revised these two paragraphs, to the form they had in the first edition (except for the addition of two sentences on the proofs). And when the revisions are compared with this passage, it is noticeable that, although the context is expanded and explained, the actual phrase that Garnett objected to is only slightly recast: "At last, the whole affair appeared in a cynical aspect to him."

It is a great pity that the notes Garnett sent and his letters to Lawrence, have not survived, for clearly his criticisms made a considerable contribution to Lawrence's development of his novel. And yet Lawrence seems often to have altered the wording to meet the criticisms, but in such a way as to reinforce his underlying intentions, even though these may have been at the root of the objection. But in some instances—as with the second of the criticisms for which full evidence survives—Garnett was able to have the last word, when he prepared the Final manuscript for the printers and pruned a tenth of it.

The survival of the manuscripts

Finally, why have certain portions of Stages II and III survived—in particular, the Fragments and the section of chapter IX—and no more? At the fourth and final stage, Lawrence wrote 437 pages out afresh onto new paper: is it plausible that he kept the first 58 pages that he had rejected from the old draft, and the twenty-three pages of chapter IX, and no others?

The survival of Stage II seems to have happened as follows: The opening 71 pages of Stage II are lost probably because, having rewritten them before falling ill in Croydon in November 1911, Lawrence left them there or threw them away at his departure. The rest of Stage II survives because he rewrote it in Eastwood and left it with his sister Ada when he went abroad. It lacked its end because he had transferred the latter pages into his third draft: and this is why he called it "the first part of Paul Morel" when he wrote home to Ada from Waldbröl on 23 May explaining for the benefit of his elder sister Emily what it was (no. 453): "Thank Emily for her letter—tell her the first part of Paul Morel is the first writing—I did it again, and have the whole here. . ."

The survival of a section of chapter IX must be related to the survival of the 'Miriam Papers'. For Lawrence had probably already rewritten chapter IX before leaving England, during the period 1-11 April when he tidied up his Stage IIIa manuscript after recovering it from Jessie Chambers with her comments written on it, and on some accompanying sheets. Then, when he

took the Stage IIIa manuscript abroad with him in May 1912, he probably also took the 'Miriam Papers' so that he could refer to them as he produced the Stage IIIb revision. He must have kept those discarded pages of chapter IX for the same reason, so that he could consult the comments Jessie had written on them.

Even so, how did the 'Miriam Papers' survive after Lawrence had finished with them? Unfortunately, Harry T. Moore, who must have seen them in private hands in or before 1964, was no more specific about their provenance than to say, referring to the section of chapter IX: ". . .it is probably a variant, for Lawrence doubtless rewrote this chapter and placed the revised version in the complete manuscript of Paul Morel B, retaining the part under discussion and putting it away with the other Miriam Papers, among which they [sic] were found after his death" (Moore 375-376; Tedlock 52). Mr. Theodore S. Bober, at the Sotheby-Parke Bernet sale of whose manuscripts in November 1980 H.R.C. bought the section of chapter IX, acquired it himself (via a Chicago dealer) from the collection of a Mr Wells, who had purchased many manuscripts direct from Frieda Lawrence.

Lawrence, then, had preserved the 'Miriam Papers' for use when writing Stage IIIb, and they were still among his papers when he died. It is possible that there were yet others, but only six items, which Frieda sold to collectors after Lawrence's death, have subsequently come to light.

But why did he preserve the first 58 pages of Stage IIIb that he rejected as he rewrote the novel, and no others?

Between finishing Stage IIIb in Icking near Munich and writing the fourth and final draft, Lawrence travelled with Frieda, largely on foot, from Icking to Lake Garda in northern Italy. As Frieda described in her memoir:[35] "We packed up our few possessions, three trunks went ahead of us to the Lago di Garda. We set off on foot, with a rucksack each. . ." The trunks went by train to Riva, at the northern end of the lake,[36] and Lawrence and Frieda, having started their trek on 5 August 1912, arrived in Riva on 5 September. On 7 September, Lawrence wrote to Garnett from Riva (no. 495): "I am glad to be settling down, to get at that novel. I am rather keen on it. I shall re-cast the first part altogether." What exactly had he unpacked from his trunk in Riva?

Back in Icking on Monday 22 July, when Lawrence received from Garnett his Stage IIIb manuscript, he wrote (no. 472): "I got Paul Morel this morning, and the list of notes from Duckworth. The latter are awfully nice and detailed. What a Trojan of energy and conscientiousness you are! I'm going to slave like a Turk at the novel—see if I won't do you credit. I begin in earnest

tomorrow—having spent the day in thought (?)". This was two weeks before he and Frieda set off on their trek on Monday 5 August.

It is probable that Lawrence did start revising his novel in those two weeks, both because of the strong desire to do so expressed in this letter, and the decision already expressed in his letter of 8 July to Garnett to "squash the first part together". I therefore propose the following hypothesis: first he read through the opening chapters and began by deleting a number of lines (see Fragment 2, pages 32-33; Fragment 5, pages 81-82, Fragment 6, pages 84-85); then he decided to rewrite some pages onto new paper, using quires of P and Q paper. He rewrote pages 1-8 on the first ten-leaf quire of P, and then he accepted pages 9-25 of Stage III and slipped them into the quire between the eighth and ninth leaves; then he carried on copying with modifications Stage III pages 26-37 onto pages 26-34 of P paper, so that when he decided to incorporate some more of Stage III pages, he had to renumber them downwards. He proceded in this fashion, revising and incorporating; and the last page of Stage III paper he drew into the new draft was page 67, renumbered 69 in the Final manuscript. I propose that he had reached Stage III page 85, the last of the Fragments, when he set off for Italy. On this page, Mrs Morel demands of the three-year-old Paul: "What are you crying for?" This scene was copied with modifications onto page 76 of the Final manuscript—and at the end of it there is a discernible change of handwriting.[37]

Lawrence apparently wrote no letter to Garnett after that of 22 July, until 4 August, the day before he left Icking, when he said (no. 476): "I am going to write Paul Morel over again—it'll take me 3 months." I suggest that this was the conclusion he had come to as a result of the labours I have described. The hypothesis is simply that he left the discarded pages (that is, the Fragments) behind—perhaps in a trunk of other possessions, probably with Frieda's sister, Else, who lived at Wolfratshausen near Icking, and whose address Lawrence still used as a forwarding address after he had left for Italy.[38] Then, in Riva, he unpacked: pages 86 to the end of Stage III together with pages 1-76 of the Final manuscript and a few blank quires of P paper. Frieda's footnote to the letter that he wrote immediately to Garnett (quoted above) suggests he may have done more reading and discussing than rewriting at first (no. 495): "I think L. quite missed the point in 'Paul Morel' . . .—he is writing P.M. again, reads bits to me and we fight like blazes over it." At any rate, he did not write any letters on P paper until 17 September, the day before they left Riva to settle for six and a half months in Gargnano, on the lakeside further south. No Q paper appears in the letters after this date: and in the

Final manuscript Q paper was used for pages 52-60 and 70, that is, during the Stage IV writing that I suggest Lawrence did in Icking.

Thus the available evidence neither proves nor disproves my hypothesis—nor does it suggest a better one. The alternative that suggests itself is that many more pages of manuscript survived, but have not yet come to light.[39]

Table 1: The Paper

• Paper A Measurements: height 209.5, width 177, thickness .011. Colour: creamy-white. Lines: 21, front and back, printed twice, at different angles, but front and back measurements are the same: space apart 9, top margin 19.5, bottom margin 14, left margin 29.5 but varying from 28 at top to 30 at bottom. Character of lines: pale blue, widish, even, but very faint. See-through features: transparency: very dense; watermarks none. Finishing features: very slight shine; flexible and strong; calendering: chain-and-wire-lines. Format: top-perforated pad, serrated top edge. Letters: nos. 256: 24 April 1911; 259: 28 April 1911; 261: 1 May 1911; 263: 7 May 1911; 264: 9 May 1911; 265: 11 May 1911; 268: 14 May 1911.
• Paper B Measurements: height 257, width 201, thickness .007. Colour: grey. Lines: 27, front only: space apart 8.5, top margin 27, bottom margin 7.5, no side margins. Character of lines: faint, fine, grey, well printed. See-through features: transparency: very transparent, shows next 2 pages; no watermark. Finishing features: matt, rather like grease-proof paper; very brittle; vertical lines on surface giving very faint crepe effect. Format: top-line pad. Letters: no. 273: 1 June 1911.
• Paper C Measurements: height 202, width across whole sheet 325, to fold 162.5, thickness .011. Colour: creamy-white. Lines: 21 front and back, one printing. Space apart varying 8-8.5, top margin sloping up from left to right 26-25, bottom margin ditto 11-13, no side margins. Character of lines, pale blue, firm, wide, well printed. See-through features, very dense, no watermark. Finishing features, slight shine; supple; calendering: chain-and-wire-lines. Format: quires of 11, 12, 12 sheets: 1: pp. 255-276, 2: pp. 277-300, 3: pp. 301-324.
• Paper D Measurements: height 250, width 203, thickness .008. Colour: grey-white. Lines: 25 front only: space apart 8.75, top margin sloping down from left to right 30.5-31.5. bottom margin ditto 11-10, no side margins. Character of lines: thick, firm even, grey-blue. See-through features, fairly transparent but not very, shows next page slightly; no watermark. Finishing features: matt; quite strong, a little brittle; mottled effect on surface. Format: top-line pad. Also used for Love Poems (Roberts E213): 'Cruelty and Love' (pp. 7, 8 out of pp. 6-8), 'Long day in Autumn', 'The Appeal', 'Repulsed', 'Dream-confused', 'Renascence', 'Dog Tired'. The dates of these poems are not known in detail; the ms. is at H.R.C.[40] Letters: nos. 281: 7 July 1911; 285: 13 July 1911; 286: [15 July 1911]; 287: [16 July 1911]; 290: 21 July 1911; 292: 24 July 1911; 303: 29 Aug. 1911; 307: 11 Sept. 1911; 313: 2 Oct. 1911; 314: 2 Oct. 1911; 315: 6 Oct. 1911; 316: 10 Oct. 1911; 317: 11 Oct. 1911; 318: 11 Oct. 1911; 321: 16 Oct. 1911; 323: 20 Oct. 1911.
• Paper e Measurements: height 250.75, width 201, thickness .007. Colour: grey-white. Lines: 24 back only: space apart 9, top margin sloping up from left to right 35-32, bottom margin ditto, 8-11, no side margins. Character of lines: grey, fine, fairly even—some

  ---

  325

  
  
  
  blobs, no breaks. See-through features: very transparent, shows next page; no watermark. Finishing features: matt; moderately brittle, high crackle; speckled on surface, forming verticle woven lines very faintly. Format: top-line pad. Also used for Love Poems: 'Cruelty and Love' (p. 6 of pp. 6-8), 'Lilies in the Fire', 'Coldness in Love', 'Reminder', 'Bei Hennef', 'Corot', 'Morning work', 'Transformations'. Letters: nos. 328: 26 Oct 1911; 329: 30 Oct 1911; 330: 3 Nov. 1911.
• Paper f Measurements: height 259, width 201, thickness .012. Colour: cream-white. Lines: 27 front only: space apart varying from 8.5-10, top margin varying from 23-25, bottom margin varying from 8-10, no side margins. See-through features: dense; watermark: Court Royal plus device: shamrock, rose, thistle spray. Finishing features: matt; supple, strong; calendering: chain-and-wire-lines. Format: top-line pad. Letters: nos. 338: 15 Nov 1911; 339: 21 Nov 1911; 341: 2 Dec 1911; 342: ?4 Dec 1911; 343: 4 Dec 1911; 345: 7 Dec 1911; 356: 9 Dec 1911; 348: 13 Dec 1911; 349: 13 Dec 1911.
• Paper g Measurements: height 251-2, width 202, thickness .007. Colour: cream-white. Lines: 24 back only: space apart 10, top margin sloping down from left to right 34-36, bottom margin ditto 10-8; no side margins. Character of lines: grey, quite even and continuous but with some blobs, though no breaks. See-through features: very transparent, shows next page; no watermark. Finishing features: matt; moderately brittle, high crackle; speckled surface. Format: top-line pad.
• Paper h Measurements: height 251.25-251.5; width 201; thickness .007. Colour: grey-white. Lines: 26 front only: space apart 9, top margin sloping up from left to right and varying 25-24, 26-25, 27-26; bottom margin ditto 10-10, 7-8, 8-9; no side margins. Character of lines: light grey, fine but unbroken, not very blobbed. See-through features: very transparent, shows next page; no watermark. Finishing features: matt; moderately brittle, high crackle; wrinkled crêpe effect from faint fine lines running length of paper. Format: probably top-line pad.
• Paper j Measurements: height 251.75-252, width 202, thickness .007. Colour: cream-white. Lines: 26 front only: space apart 9, top margin sloping down from left to right 24-26, bottom margin ditto 11-9, no side margins. Character of lines: grey, fairly firm but fine. See-through features: very transparent, shows next page; no watermark. Finishing features: matt; moderately brittle, high crackle; slightly mottled. Format: probably top-line pad. Letters: nos. 396: 1 March 1912; 397: 4 March 1912; ?412: 5 April 1912; ?415: 11 April 1912; ?418: 19 April 1912.
• Paper k Measurements: height 250, width 203.75-204, thickness .007 Colour: creamywhite. Lines: 24 front only: space apart 9, top margin sloping up from left to right 32.25-29.75, bottom margin ditto 11-12.75, no side margins. Character of lines: grey, fairly firm, but fine. See-through features: very transparent, shows next page; no watermark. Finishing features: matt; moderately brittle, high crackle; very light mottled effect. Format: possibly sheets, not much evidence. Letters: nos. 399: 6 March 1912; 400: 8 March 1912.
• Paper l Measurements: height 251.5 at left, 252 at right, width 203-4, thickness .007. Colour: grey-white. Lines: 26 front only: space apart 8.5, top margin sloping down from left to right 23.5-26, bottom margin ditto 10-8.5, no side margins. Character of lines: fine, grey, marginally more uneven than other lined papers: g, h, j, k. See-through features: very transparent, shows next page; no watermark. Finishing features: matt; moderately brittle, high crackle; faint mottled effect. Format: top-line pad. Letters: no. ?407: 27 March 1912.
• Paper m Measurements: height 270, width 212, thickness .009. Colour: pale blue. Lines: 24 front, 27 back. Space apart: front and back 9, top margin front sloping up from left to right 51.5-50, back ditto 24.5-23, bottom margin ditto front and back 11.5-13, no side margins. Character of lines: grey, fine firm both sides. See-through features: dense, not transparent; no watermark. Finishing features: slight shine, smooth; strong, good quality, quite supple; marked vertical grain, light horizontal grain. Format: probably sheets. Letters: nos. 445: 16 May 1912; 446: 16 May 1912; 447: 17 May 1912; 452: 21 May 1912.

---

326




• Paper n Measurements: height 284, width 226, thickness .006. Colour: grey-white. Lines: none. See-through features: extremely transparent, shows next two pages; no watermark. Finishing features: like grease-proof paper; very high crackle, buckled and creased for an inch on each side. Format: sheets. Also used for Once (Roberts E296), written in Icking sometime between 2 and 29 June 1912; 13 pages. Letters: nos. 463: 29 June 1912; 464: 3 July 1912; 467: [8] July 1912; 471: 18 July 1912. The letters and the pages of IIIb both written in brown ink; the Ms. revised in blue-black.
• Paper P Measurements: height 287 at left and 288 at right, width 221, thickness .00725. Colour: cream. Lines: none. See-through features: slightly transparent, dimly shows next page; no watermark. Finishing features: shiny, highly glazed; crisp but not brittle nor limp, moderately high crackle; very strong 'linen' lines—lines vertical and horizontal. Format: quires of up to five sheets; torn by printer. Letters: nos. 500: [17 Sept. 1912]; 503: 3 Oct 1912; 505: 5 Oct 1912.
• Paper Q Measurements: height 289.5 at left, 290 at right; width varying to tear (torn sheets) 223-225; thickness .008. Colour: creamy but not as yellow as P. Lines: none. See-through features: very slightly transparent, faintly shows that there is writing on the next page; no watermark. Finishing features: slight shine; supple, not brittle; moderately high crackle; calendering: marked chain-and-wire-lines, with chain lines 25 apart, and wire lines 1.5 apart. Format: one quire of five sheets, ten leaves; torn by printer.
• Paper R Measurements: height 339.5, width to tear (torn sheets) 211, thickness .0075. Colour: cream-white. Lines: none. See-through features: fairly transparent, shows next page; no watermark. Finishing features: matt; not brittle, flexible and some pages quite flimsy; dotted effect over surface. Format: quires of very variable sizes; torn by printer.
• Paper S Measurements: height 340, width varying to tear (torn sheets) 210-211, thickness almost .008. Colour: orangey-white: orange hairs or fluff or flecks on surface. Lines: none. See-through features: fairly dense; no watermark. Finishing features: shiny, highly glazed; firm but not stiff, flexible but strong, moderate crackle; very faint dotted diagonal marks on surface. Where torn the paper frays badly. Format: quire, one gathering of 8 sheets, 16 leaves: pp. 294, 305-319. Also used for Christs in the Tirol (Roberts E81.5). Letters: nos. 489: 30 Aug 1912; 493: 2 Sept 1912; 495: 7 Sept 1912; 496: 11 Sept 1912; 498: 16 Sept 1912; 560: 22 March 1913; 564: 5 April 1913.
• Paper T Measurements: height 297, width 228, thickness .007. Colour: grey. Lines: none. See-through features: moderately transparent, shows next page; no watermark. Finishing features: slight shine; quite stiff, high-pitched crackle; surface has strong vertical lines, making mottled-ribbed effect at a slope. Format: twelve quires of four to six sheets. Letters: nos. 507: 15 Oct 1912; 509: 29 Oct 1912; 511: 7 Nov 1912; 516: 19 Nov 1912.
• Paper U These are mixed quires of two slightly different papers: U(i) and U(ii). Measurements: height 310, width to tear (torn sheets) varying 208-211, thickness varies from page to page between .0075 and .009 but mostly .008 and .0085. Colour: U(i) browny cream, very flecked, seems dirty; U(ii) creamy, unflecked, seems cleaner. Lines: 25 front and back, one printing. Variations in margin measurements arise from the fact of torn folio sheets: space apart 10, top margin front 34 back 34.5, bottom margin 36, left margin front 28 back 48, right margin front 47 back 28.5. Character of lines: U(i) very fine and faint, grey, not printed right to outer edge; U(ii) grey, fine, but clear and printed to edge. Character of margins same as that of lines. See-through features: very dense; no watermark. Finishing features: U(i) moderately shiny and smooth, U(ii) more shiny and smooth; U(i) fairly firm but rather supple without being flimsy, U(ii) similar but firmer. All pages torn at left and very frayed. Format: five quires of 5 or 6 sheets and some odd leaves. Also used for The Fight for Barbara (Roberts E130). Letters: nos. 510: 30 Oct 1912; 513: 14 Nov 1912.

---

327

Table 2: The Mss

• The locations given in this appendix are as follows:
• Paul Morel Ms. H.R.C., Austin (Roberts E373.d).
• Fragments Bancroft Library, Berkeley (Roberts E373.a).
• Final Ms. Bancroft Library, Berkeley (Roberts E373.e).
• Section of Ch. IX H.R.C., Austin (Roberts E373.b).

Stage II Written 13 March to July or October 1911

• pp. 72-165 A paper, used in letters 24 April to 14 May 1911. Now located: Paul Morel Ms.
• p. 166 Lost
• pp. 167-254 B paper, used in letters 1 June 1911. Paul Morel Ms.
• pp. 255-324 C paper, not found in letters. Paul Morel Ms.
• pp. 325-353 D paper, used in letters 7 July to 20 Oct 1911. Now located: pp. 325-335, 353 Paul Morel Ms. pp. 337-8, 341-351 (347-9 in error written as 247-9) Final Ms. pp. 372-3, 377-387

Stage III Written (IIIa) 3 to 19 Nov 1911, between 9 and 23 Feb to c.25 March 1912; corrected 1-11 April 1912; revised (IIIb) 16 May to between 2 and 9 June 1912. A reconstruction of the surviving segments in the chronological order of the main paper-use; many pages are now lost.

• False start: pp. [1-7] unnumbered and unfinished, on e paper, used in letters 26 Oct to 3 Nov 1911. Now located: Paul Morel Ms.

IIIA

• pp. 1-74 f paper, used in letters 15 Nov to 13 Dec 1911. Now located: pp. 1-8, 26-36, 44, 46-58, 68-74. Fragments. pp. 9-25, 37-43 (renumbered IV 35-41), 59-61 (IV 61-3), 63-67 (IV 65-69) Final Ms.
• pp. 75-85, 190-192 g paper, not found in letters. Now located: pp. 75-85 Fragments. pp. 190-2 (renumbered IV 202-4) Final Ms. The displacement of these three g pages in the paper sequence in Stage III must be because they became separated from the batch of g paper, and Lawrence decided to use them up after he had already begun using h.
• pp. 182-189, 204-226 h paper, not found in letters. Now located: pp. 182-9 (renumbered IV 194-201) Final Ms. pp. 204-226 Section of Ch. IX
• pp. 272-281 j paper, used in letters 1,4 March and possibly 5-19 April 1912. Now located: Final Ms. pp. 295-304
• pp. 309-318 k paper, used in letters 6-8 March 1912. Now located: Final Ms: Separated into two parts and reversed: pp. 317-318 (renumbered IV 345-346) and pp. 309-315 [misnumbered 309-312, 314-315 by Lawrence] (renumbered IV 350-355). This caused Lawrence at first to mis-number the three T pages in between 347-349, IV according to the Stage III numbering,
• pp. 328-331 l paper, perhaps used in letters 27 March 1912. Now located: Final Ms. pp. 368-371
• pp. 332-333, 337-347 D paper brought in from Stage II and renumbered. (Original numbering in Stage II: 337-8, 341-351) Now located: Final Ms. pp. 372-3, 377-387

IIIB

• pp. 353-5, 358-362, 370-373 m paper, used in letters 16-21 May 1912. Now located: Final Ms. pp. 397-9, 402-6, 414-7
• pp. 45, 46a-49a, 62 n paper, used in letters 29 June to 18 July 1912. Also used for Once, written between 2 & 29 June. Now located: pp. 45-49a Fragments pp. 62 Final Ms. p. 64

  ---

  328

  
  
  
  Also in Fragments on n paper two unfinished and therefore probably immediately rejected pages, numbered 68, 69, bearing an alternative end to Chapter II.

Stage IV Written [?23 July to ?4 August and] 7 September to 19 November 1912

• pp. 1-88 P and Q paper (Sequence 71-97 misnumbered by Lawrence, p. 88 renumbered 91 by Garnett.) P and Q paper are similar: both are in quires, and Lawrence alternated the quires:
• pp. 1-51 P Paper, used in letters 17 Sept to 5 Oct 1912. Three quires interspersed with Stage III pages (in IV numbering): f 9-25, 35-41
• pp. 52-70 Q paper, not found in letters. One quire pp. 52-60, 70, interspersed with Stage III pages (in IV numbering): f 61-63, n 64, f 65-69.
• pp. 71-88 (Garnett's 91) P paper, as above. Two quires and some miscellaneous leaves.
• pp. 89-293 R paper, not found in letters. Odd-sized quires, varying from 4 to 20 leaves, interrupted by only one sequence of Stage III pages (in IV numbering): h 194-201, g 202-4.
• pp. 294-319 S paper used variously: 1. Letters, 30 August to 2 September 1912; 2. Christs in the Tirol 2-6 September 1912 (1 six-leaf quire); 3. Letters, 7-16 September 1912; 4. Sons and Lovers sometime between 5 and 15 October 1912 (1 sixteen-leaf quire); 5. Letters, 22 March and 5 April 1913. Interspersed with Stage III pages (in IV numbering): j 295-304.
• pp. 320-477 T paper, used in letters 15 Oct to 19 Nov 1912. Twelve quires of 8, 10, and 12 leaves, interspersed with Stage III pages (in IV numbering): k 345-6, 350-355, l 368-371, D 372-3, 377-387, m 397-9, 402-6, 414-7, 419-420.
• pp. 478-540 U paper, used in letters 30 Oct, 14 Nov 1912. Also used for The Fight for Barbara, written 27, 28, 29 October 1912: 55 pages in 6, 8 and 10-leaf quires and separate leaves. Used in Sons and Lovers in six quires of miscellaneous sizes, uninterrupted to the end of the novel: finished 19 November 1912.

Notes