This page contains links to searchable scanned PDF images of three early printings of To the Lighthouse and to PDF documents containing the texts of those editions, extracted from the scanned images. It also includes scanned images of the proofs marked for the American edition, and the text of three extensive passages deleted from the marked proofs in order to reduce the page count of the printed book. This page also includes a lightly edited text of the November 1926 typescript of "Time Passes" made by the author for separate translation into French; a note on possible corrections in the first French translation; and notes on existing editions of the novel.
This site also offers similar pages about Mrs. Dalloway and The Waves and other items.
The three printings of the novel that have textual authority are: the first American edition and the first British edition, both from 1927, and the 1930 fourth impression of the first British edition, published as part of the Uniform Edition of the Works of Virginia Woolf, with revisions evidently made by Virginia Woolf. ("Textual authority" means that these printings derive directly from the hand of the author.)
The American edition, although published on the same day as the British edition, represents an earlier state of the text. The American text is based on proof corrections made by the author before she made her final proof corrections for the British text.
Scans and texts of the three authoritative printings are linked here:
All other editions published during the author's lifetime (those by The Albatross in 1932, for sale on the European continent, and by Everyman's Library in 1938, both based, with errors, on the 1930 Uniform Edition text), and all editions published after the author's death in 1941, have no independent textual authority. (An incorrect description of the text of The Albatross edition at woolfonline.com was corrected in July 2025.)
The Albatross edition is notably defective, with four paragraphs printed out of sequence; the paragraphs from "The children are disgraceful" through "a good deal in what he says" (Hogarth editions, pages 145-147) are printed between "indeed it was time" and "Light the candles" (Hogarth page 149), probably because the galley containing those paragraphs was misarranged by The Albatross's printers.
Vastly superior scanned images of the first American edition, of the first British edition, and of the 1932 photo-offset reprint of the 1930 impression, together with scanned images of the author's notebooks, typescripts, corrected proofs, and much else, may be found at woolfonline.com (but see the notes below on that site's account of the printing and production of the novel).
Virginia Woolf made sixteen changes to the February 1930 fourth impression, for the "Uniform Edition of the Works of Virginia Woolf" (page numbers refer to the scanned PDFs):
The third impression (May 1928) had earlier introduced a printing error that persists in the fourth impression: Page 216, line 7: the comma dropped out of the plate at the end of the line, after "upstairs now", leaving a narrow blank space between "now" and the right margin.
The fourth impression added a further printing error: Page 119, line 17: "power" was misprinted "dower", the first letter having come loose from the plate and restored upside-down.
The fourth impression (with the exception of the two errors noted above) is the most accurate text, but probably requires eleven further emendations, all (unless noted) suggested or applied (some silently) by Stella McNichol in her 1992 Penguin and Macmillan editions (page numbers refer to the scanned PDFs):
About the three emendations of "an earwig" to "a earwig": Virginia Woolf presumably marked these changes in the lost British proofs (as she marked two on the Harcourt proofs), only to have them silently rejected by the printer. (As noted below, she may not have noticed one instance of "An earwig" in the Harcourt proofs because "An" occurred at the end of a line of type and "earwig" at the start of the next.)
Stella McNichol's Penguin and Macmillan editions (see below) seem to be the only ones that take explicit notice of the changes in the 1930 impression. However, in the Penguin edition, she suggests in her notes (and in some instances applies to her text) a few further emendations that seem either mistaken or unnecessary.
Virginia Woolf's draft manuscript correctly numbers the sections of "The Lighthouse," but the British editions misnumber all the sections after the first; the section numbered 3 is the second section, the section numbered 4 is the third, etc. The American edition corrects the numbering (and uses roman numerals, not arabic). I believe that this error ought not to be corrected, as the section numbers are those that Virginia Woolf saw on the page, and they may, possibly, have had particular significance for her. (The 1932 Albatross Edition corrects the misnumbering. The 1938 Everyman's Library edition preserves the erroneous numbering of sections 3 to the end, but creates a spurious section 2 by dividing the first section into two, starting its section 2 with "Suddenly Mr. Ramsay raised his head".)
The sixteen differences between the first British edition and the 1930 Uniform Edition may be seen in this PDF document that displays the variants in different colors (blue and struck through for the first edition, red and underlined for the Uniform Edition).
The many differences between the first American edition and the first British edition may be seen in this PDF document that displays the variants in different colors (blue and struck through for the American edition, red and underlined for the British).
The American edition has a large number of printing errors where the compositor misread either Virginia Woolf's markings or the printed text of the proofs, and it has a few authorial errors where Virginia Woolf's markings were incorrect or incomplete. Surprisingly, no edition based on the Harcourt text attempts to correct these errors, although Susan Dick's edition of the Harcourt proofs (without Harcourt's editorial changes) avoids most of them. The page numbers below refer to the scans of the American edition (and I have not attempted to record all the commas inserted by the compositor that do not occur in the proof or British text).
The 1938 Everyman edition has three verbal variants that seem to be compositor's errors (page numbers refer to the scans of the Uniform Edition):
The text recorded in Virginia Woolf: To the Lighthouse: The Original Holograph Draft, transcribed and edited by Susan Dick (1983), is an earlier state than any of the printed versions. Scans of the draft may be found on the woolfonline.com web site.
I am grateful to Mark Hussey, Stephen Barkway, and Stuart N. Clarke for indispensable help in preparing this page.
In November 1926 Virginia Woolf made a separate typescript of "Time Passes" for translation into French by Charles Mauron, published in Commerce, cahier 10, hiver 1926 (probably in January 1927).
A lightly edited text of the typescript may be found here. I have corrected a minor typing error ("tress" for "trees" on p. 6, line 5); added a comma between "step" and "window" on p. 23, 3 lines up; and have added a possessive apostrophe in "cliff's" on p. 30, last line. On p. 11, line 4, Virginia Woolf inserted the eight words starting "leaning her" following the full stop after "it all"; I have changed that full stop to a comma.
Scanned images of the typescript (and of the translated text in Commerce) may be found at woolfonline.com. A transcript of the text was first published by James M. Haule in Twentieth Century Literature, Autumn 1983. Another transcript was published in Susan Dick's 1992 Shakespeare Head Edition of the novel (see below). Both transcripts (unlike the edited text provided here) indicate the author's deletions, insertions, and revisions.
One full and one partial set of proofs, marked for Harcourt, Brace to use when setting the American edition in type, are in the Mortimer Rare Book Room of the William Allan Neilson Library of Smith College. These are reproduced in the form of separate images on woolfonline.com and, on this site, as a PDF of the full set and a PDF of the partial set.
The Harcourt proofs comprise a revised full set of gatherings A through U (omitting J; pages numbered 1–321 [322-324]), and a partial set of gatherings S through U (pages numbered 273–321 {322-324]).
Virginia Woolf first sent Donald Brace a first full set of proofs (comprising the copies of gatherings A through T now included in the revised full set, and the copy of gathering U now included in the partial set described above). She then sent second copies of gatherings S and T to Brace on 1 March 1927, containing additional corrections on pages 286-290 that were then copied by a Harcout editor to the full set; these second copies of gatherings S and T did not replace the first set, but were used only to make corrections in it. The second copies are now part of the partial set described above. At this point, she seems to have sent second copies of only gatherings S and T, which include pages 286-90.
Some days later, in an undated letter (received 16 March 1927), she sent a second copy of gathering U, describing it as "the final pages of To the Lighthouse" in which she has "made some further corrections". This second copy of gathering U is now part of the revised full set described above, where it replaces the version orringinally sent in the first full set; the earlier version was then transferred to the partial set, together with the revised copies of gatherings S and T.
The partial set is marked on its first page by an editor: “additional corrections on pp. 286–91 transferred to first set”. Gathering U in the full set is marked by an editor “305–322 New Copy recd 3/16/27”; the earlier set of gathering U, now in the partial set, is marked by an editor “305–322 Old Copy—new copy recd 3/16/27”. As noted by the American editor, some corrections in gatherings S and T in the partial set were transferred to the full set. The “New Copy” of gathering U was physically placed at the end of the full set, and the “Old Copy” was moved to the partial set, where it follows the newly-arrived gatherings S and T.
The full set contains, in addition to the only copies of gatherings A through R: the older copies of gatherings S and T, with markings transferred by an editor from the newer copies of the same gatherings (which are preserved in the PDF of the partial set); and the newer copy of gathering U. The full set is the set that Harcourt used for typesetting.
All the gatherings have been opened and cut, producing single leaves, except for the newer copies of gatherings S and T, which remain folded and gathered; the images at woolfonline.com show the unfolded facing pages, so that (for example), pages 274 and 287 are shown on the same image, rather than presenting the pages in actual numerical order. The PDF file created for this site splits the images of the facing pages in the newer copies of gatherings S and T, showing the pages in correct numerical order.
Gatherings A through T in the proofs were standard sixteen-page gatherings (four sheets, each folded to make four pages). Gathering U a comprised single sheet (four pages) folded around a sixteen-page gathering signed U 2; the last two pages of this twenty-page gathering were blank. To avoid the expense required by printing and binding the additional four-page sheet, Virginia Woolf deleted text amounting to two pages, permitting the final text to end with a standard sixteen-page gathering. Three extensive passages deleted in the marked proofs in order to reduce the page count of the printed book are transcribed here. The deletions in the American edition are slightly more extensive.
Notes: The account of the proofs that appeared at woolfonline.com until July 2025, and repeated by some editors and critics, has now been corrected. Contrary to the account posted before July 2025, there was no "single loose leaf at the end"; the full text, as it existed before the author made her revisions, would not have required "a signature V or W at the foot of page 321." The printer had already made room for the full text in gatherings U and U 2, and no additional gathering was needed. (And the author of the notes at woolfonline.com was evidently unaware that printers did not use "V or W" for signatures, but followed U with X.)
Also: two images on the woolfonline.com site that until July 2025 were wrongly placed in the site's galleries have now been removed entirely, rather than being moved to their correct positions. The two removed images may be seen in the PDF on this site: they are the two last blank pages in the full set of proofs.
These marked proofs, like all of Virginia Woolf's manuscript and unpublished works, are protected by copyright in the United Kingdom (not elsewhere) through 2039. They are posted here by permission of The Society of Authors as the Literary Representative of the Estate of Virginia Woolf. The Estate further specifies: “Users of the website may make copies of the Works under general copyright exceptions, but cannot further copy, share or adapt the Works. This shall only be for non-commercial purposes and the copyright holder shall be credited.” These restrictions apply only in those places where the proofs are protected by copyright.
The first French translation of the novel, by M[aurice] Lenoire, was published as La Promenade au Phare (Paris: Librarie Stock, Delamain & Bouttelleau, 1929). A scanned PDF is posted here. It seems to contain a correction (perhaps two corrections) possibly made by Virginia Woolf. It may contain others, but these are the only ones that I have noticed.
The first correction resolves a major textual crux in the novel, which, as noted briefly above, occurs in "Time Passes", in the last paragraph of section 3. The first Hogarth edition reads:
[Mr. Ramsay stumbling along a passage stretched his arms out one dark morning, but Mrs. Ramsay having died rather suddenly the night before he stretched his arms out. They remained empty.]
This is clearly defective. The 1930 Uniform Edition adds two commas:
[Mr. Ramsay stumbling along a passage stretched his arms out one dark morning, but, Mrs. Ramsay having died rather suddenly the night before, he stretched his arms out. They remained empty.]
This is less obviously defective, but still not quite right. Stella McNichol's proposed (and convincing) emendation makes the minimal change of moving the second comma to the location occupied in the printed text by the full stop:
[Mr. Ramsay stumbling along a passage stretched his arms out one dark morning, but, Mrs. Ramsay having died rather suddenly the night before he stretched his arms out, they remained empty.]
Presumably the changes in the 1930 text were made in a letter sent to the printer by Leonard Woolf, and the printer misinterpreted the instructions that ought to have produced the text proposed by Stella McNichol.
Stella McNichol's emendation gets strong support from the French text:
(Mr. Ramsay, trébuchant le long d'un couloir, étendit les bras, un matin obscur. Mais Mrs. Ramsay étant morte assez soudainement la veille au soir, ils restèrent vides.)
It seems likely that the English original that Maurice Lanoire translated had been corrected by Virginia Woolf to match, more or less exactly, Stella McNichol's emended version. (The first two commas, however, seem likely to have been added by the translator without prompting from the author.) M. Lanoire had written to Virginia Woolf on 4 June 1928 asking for help with some phrases in the novel, and reporting that he had sent separately the first fifty pages of his version. He wrote in his letter: "You will notice that in some instances I have had to break up the flow, the admirable flow, of your sentences or periods. The French language does not possess the plasticity of the English and its roll cannot always have the same length. I have however tried to maintain, as much as possible, the r[h]ythm and the swing of your style." Virginia Woolf's lost reply to his letter may well have included her correction to the paragraph about Mr. Ramsay.
The corrected version that she may have sent to M. Lanoire seems likely to have been a single sentence, as in Stella McNichol's emended version. The French version, in the manner that M. Lanoire suggested in his letter, breaks up the flow of the sentence by dividing into two where it could logically be broken - but not at the erroneous sentence break in the original Hogarth Press text ("his arms out. They remained").
Note: The 1931 German translation by Karl Lerbs, Die Fahrt zum Leuchtturm (Leipzig: Insel-Verlag) follows the defective text in the first Hogarth edition. The 1934 Italian translation by Giulia Celenza, Gita al Faro (Milano: Fratelli Treves Editori) retains the erroneous sentence break but, like the French translation, adds commas near the beginning of the passage; the commas seem likely to have been added by the translator without prompting from the author (pp. 177-78):
(Il signor Ramsay, incespicando per un corridoio, tese le braccia in un mattino oscuro; ma siccome la signora Ramsay era morta piuttosto improvvisamente la notte avanti, egli tese le braccia indarno. Queste rimasero vuote.)
The second possible correction in the French translation occurs where the British editions have "in one of those habitual currents which after a certain time forms experience in the mind" (p. 246), instead of "in one of those habitual currents which after a certain time experience forms in the mind" (similar to the reading of the American text); the British compositor probably misread a correction in the proof from "time wisdom forms, so that". The French translation (p. 189) renders what is clearly the correct version, either because Virginia Woolf made the correction or because the translator inferred the correct reading by interpreting the syntax of the English text as if it had been written in French:
dans un de ces courants d'habitude que l'expérience forme au bout d'un certain temps dans l'esprit
No existing edition is based, as I believe an edition ought to be based, on the text of the first Hogarth edition, altered with the sixteen revisions that Virginia Woolf made to the 1930 impression (but not the two printing errors introduced in the 1928 and 1930 impressions), and the eleven emendations described above. An edition that records variant readings also should, I believe, present its data in readable form, not merely in skeletal form as a list.
Many existing editions of To the Lighthouse present a text prepared by an editor, not merely reprinted without explanation from an earlier version; these are described below. Some other editions, not listed below, include explanatory notes but no notes on the text. Editions that identify themselves as using the first-edition text, but in fact use a version of the text from 1930 or later (e.g. the 2012 Alma Classics edition), may be identified by any of the sixteen 1930 variants listed above. At least one edition, the 1990 Vintage (UK) edition frequently reprinted with introductions by various writers, claims to be based on the 1927 text but combines readings from the 1927 and 1930 texts.
The anonymously-edited 1990 Hogarth Press "Definitive Collected Edition" follows the text of the first British edition. It includes a partly inaccurate list of substantive variants between the first British and first American editions, based on the list in A Concordance to To the Lighthouse by Virginia Woolf, by James M. Haule and Philip H. Smith, Jr. (1983), who used a 1977 reissue of the Hogarth Press edition and a 1964 reprint of the American text, mistaking the printing errors in these editions for variants in the original texts.
Susan Dick's 1992 Shakespeare Head Press edition (Blackwell Publishers) is based (for weakly defended reasons) on the marked proofs of the American edition, but is the version of that edition that most closely reflects authorial choices, silently correcting some printer's errors in the Harcourt text and correctly untangling confusions in the proofs. It includes an extensive (but not entirely accurate) list of variants from the unmarked proofs (including long passages cut before publication) and from the first British edition. It makes intelligently-chosen emendations, some of them matching the author's later revisions (although the edition does not note the existence of those revisions). It also includes the full text of the early typescript of "Time Passes" that Virginia Woolf prepared for translation into French.
Note: The list of variant readings in Susan Dick's edition omits the variants on p. 266, 6 and 4 lines up: the Harcourt proofs, followed by the American text, are marked to delete the sentence "Something violent." and to replace "while he spoke. He spoke indignant, jealous words, abusing her" with "while he spoke something violent, abusing her"; these changes are not made in the British text.
Margaret Drabble's 1992 Oxford World's Classics edition describes itself as using "the Hogarth Press text, first published in 1927"; its text is in fact that of the 1930 fourth impression (the existence of which is not mentioned by Drabble), or a later reprint. It includes a selective list of variants between the first British and first American editions.
Stella McNichol’s 1992 Penguin Twentieth-Century Classics edition, with an introduction by Hermione Lee, is based on the 1927 first edition, with some but not all of Virginia Woolf’s later changes; with well-considered emendations (and, as noted above, a few mistaken ones); and with notes about some other variant readings. This, and the same editor's Macmillan volume noted below, seem to be the only editions that take into account the changes in the 1930 fourth impression. The Penguin Twentieth-Century Classics edition has often been reprinted and reset (with formatting errors not present in the 1992 edition) under other Penguin imprints. It includes a selective list of variants between the first British and first American editions.
Stella McNichol’s 1992 edition of the novel in a Macmillan “student compendium,” Collected Novels of Virginia Woolf: Mrs. Dalloway, To the Lighthouse, The Waves, has a text based on the 1930 fourth impression, with a few regularizing emendations.
Sandra Kemp's 1994 Routledge English Texts edition describes its text as "based on that of the original Hogarth Press edition"; its text seems to combine readings from the first British edition and the 1930 fourth impression (which is not mentioned by Kemp).
Mark Hussey's 2005 annotated edition uses the American text (as copyright required at the time), with a few emendations; the textual notes discuss some textual matters.
David Bradshaw's 2006 Oxford World's Classics edition follows the text of the first British edition with minor house-style variants. Its note on the text suggests that readings in Margaret Drabble's 1992 edition were errors, but those readings were in fact derived from the 1930 fourth impression (which is not mentioned by Bradshaw). It includes a selective list of variants between the first British and first American editions.
An anonymously-edited 2021 Renard Press (London) edition has a note claiming that its text "is based on the text of the first edition," with "spelling, punctuation and grammar ... silently corrected." The text is in fact based on the corrected fourth impression of the first edition or a later reprint.
Margaret Homans's 2023 Norton Critical Edition uses the text of the American edition (without stating any reason to do so), with minor (and some mistaken or mistakenly reported) emendations. It includes an extensive but selective list of variants among the unmarked proofs, the first British, and the first American editions, and some variants from the typescript of "Time Passes" that Virginia Woolf prepared for translation into French. Its detailed chronology is by far the most useful history of the composition and revision of the book.
All editions first published in North America (including those not listed here because they make no claims about their text), and the one edition edited by a North American scholar for publication in the UK (Susan Dick's edition), are based on the American text.
American vs. British texts: A common myth among American editors of Virginia Woolf's novels states that no conclusive evidence favors either the American or the British edition as Virginia Woolf's more-considered or later-considered text. This defies both reality and common sense. In order for the American and British editions to be published on the same day, as Virginia Woolf had arranged for most of her novels, she was obliged to send proofs for the American edition at least two weeks before she was obliged to return proof for the British edition. One of those two weeks was required by shipborne transatlantic mail; the other of those weeks was required for the full resetting the text by an American printer, as was then required by American copyright law. For the British edition, she could return the marked proofs to Edinburgh via overnight mail for the printers to make minor corrections requiring at most a few hours of labor. Absolutely no plausible reason exists for preferring the earlier American texts to the later British ones.
I prepared these scanned images by using a Czur ET-24 Pro book scanner to make digital copies of three versions of the text: the first and fourth ("Uniform Edition") Hogarth Press impressions from 1927 and 1930; and the Harcourt (American) first edition from 1927. The less-than-perfect quality of the scanned images is the result of (1) my incompetence, (2) the relatively low-priced scanner that I used, and (3) the cheap, battered copies of the original editions that I could afford to buy.
I corrected the scanned output by proofreading all four versions in the OCR editor features of ABBYY FineReader and Adobe Acrobat Pro; then, using those applications and Microsoft Word, I compared the scanned texts to each other in order to identify variant readings and remove any remaining scanning errors.
Edward Mendelson (edward [dot] mendelson [at] columbia [dot] edu)