"Six medical students under my direction will write Paradise Lost except 100 lines." This claim of Stephen's does not make it into Ulysses. John Eglinton's jibe, however, "asked with elder's gall" at U 9.18–19 makes it clear that Stephen has at some time alluded to his "six brave medicals". The material absent from Ulysses but implied by Eglinton's question has been known to readers since it appeared in Herbert Gorman's 1939 biography of Joyce.2 There Gorman placed it alongside the Pola 1904 portions of a notebook Joyce had begun in Paris in 1903. Yet when Scholes and Kain reprinted the material quoted by Gorman in The Workshop of Daedalus they saw fit to divide the notebook up into distinct Paris and Pola notebooks.3 Quite why they set keenly to the splitting of such hairs and made distinctions where none existed is unclear. Gorman explicitly states that the notebook Joyce had begun in Paris was continued in Pola. Joyce recorded his scraps and notes for Portrait, Gorman says, in the same "early cahier".4 The allusion to a cahier is not only indicative of Gorman's godwottery; when the notebook turned up in 2002 it bore the imprint of a French stationer.5 This mathematics notebook with its red-brown cover, designated MS 36639/2A in the National Library of Ireland collection list, dates from 1903–04 and contains budgets, literary excerpts and notes alongside handwritten versions of the famous passages on aesthetics (i.e. the contents of the so-called Paris Notebook and Pola Notebook). Stephen's Miltonian pronouncement appears on p. [22v] of /2A, alongside two similar statements for "S. D.", which were given by Gorman and Scholes and Kain.6 The unearthing of the Léon cache not only dissolves the Paris-Pola confusion but it is also shining light on Joyce's "dark materials" – his notes for Ulysses.

Reminiscences of the composition of Ulysses abound with references to Joyce making notes on little writing blocks; on menus, bus tickets and in the margins of newspapers; on stray bit of paper or the backs of advertisements; and, on one occasion, on the cuff of a shirtsleeve. This heterogeneous array of first-order notes, Joyce's "scribblings", has not survived. Instead, the extant body of Ulysses notes includes several notebooks and, in the British Library, loose sheets of notes, more commonly termed 'notesheets', covered with lists of words and short phrases, repositories of the original 'off-the-cuff' notes. Prior to the acquisition of the Léon cache by the National Library of Ireland, only two notebooks were known to be extant: Buffalo MS VIII.A.5, a commonplace notebook compiled in the Zentralbibliothek while Joyce was living at 38 Universitätstrasse Zürich in 1918, and "the deteriorated notebook" Buffalo MS V.A.2, a repository of late notes for most episodes of Ulysses, intended for typescripts and placards.7 The C-series of notebooks at Buffalo, comprising Mme France Raphael's transcriptions, suggested that there were at least two other Ulysses notebooks at one time: Spielberg's MS VI.D.4 and Rose and O'Hanlon's reconstructed "Lost Notebook", MS VI.D.7.8

The acquisition of the Joyce Papers 2002 brought the number of Ulysses notebooks known to be extant to six. The Léon cache includes the so-called "Subject Notebook", NLI MS 36639/3, and three notebooks designated 36639/4, /5A and /5B. The Subject Notebook is a mathematics exercise book with pink-red covers. Apparently it was bought and begun when Joyce was convalescing in Locarno in the winter of 1917, and it contains notes under such headings as "Simon", "Leopold", "Books" and "Recipes". NLI MS 36639/4 is a twenty-four page notebook with pale blue covers containing pages labelled with the episode titles in order but skipping "Sirens", one page labelled "Eventuali", five pages labelled "Penelope", and one "Circe". NLI MS 36639/5A is a notebook with "[p]urple-grey marbled outer cover; grey-green plain inner cover" and contains sixty unnumbered pages labelled firstly on the rectos with the titles of the episodes in order of the Odyssey, the Telemachiad, and the Nostos, then with a seemingly random order (which includes two pages labelled "Eventuali"). A second pale-blue, smaller notebook, NLI MS 36639/5B, contains notes again under the episode titles (in a seemingly random order) and two chronologies of Bloom's life.9

In this paper I examine the expanded genetic dossier of Ulysses note repositories. The bibliography of the recently surfaced notebooks will be familiar to many, but only recently have the relationships between these documents and the preparatory materials that were long known to be extant begun to be established. I examine the connection between the National Library's notebooks and Buffalo MS VI.C.7, and speculate about the implications for Rose and O'Hanlon's reconstituted MS VI.D.4. I also trace Joyce's use of the notebooks in compiling notesheets BL "Circe" 16 and 17, the two sides of BL Add. MS 49975/fol. 20.

The harlequined pages of the notebooks contain entries that fall into the usual lists of words and short phrases, repositories of earlier "scribblings", familiar from the notesheets and Buffalo MS V.A.2. A few pages of the smaller of the two pale-blue notebooks, NLI MS 36639/5B, contain notes in Italian in Lucia's hand. This collaboration between Joyce père and fille raises the issue of the immediate provenance of the entries; are they first-order notes – "the written symbols of the languid lights which occasionally flashed across [Joyce's] soul" (LI: 154) – and was Lucia writing from sore-eyed Joyce's dictation? Or are they second-order remnants of earlier Italian "scribblings" in Joyce's hand here transcribed by Lucia? Do the entries borrow from Italian sources to which Joyce had directed Lucia? None of these explanations is entirely satisfactory but such questions imply the broader inquiry into the intertextual origin of all the notebooks' entries, an inquiry vitiated by the lack of any identifying tags to go with the entries. The earlier 1903–04 notebook acquired by the National Library, NLI MS 36639/2A, contains, alongside its budgets and passages on aesthetics, excerpts copied from Ben Jonson, Aristotle, and Yeats, among others. Whereas these latter quotations are acknowledged as such with authors and titles of works listed, notes on pages labelled "For Stephen Hero" and "For Dubliners" (and, indeed, in the four Ulysses notebooks) have no such indications as to their provenance. Even reading notes cannot be distinguished from creative acts.

The odyssey undertaken by these "dark materials" across the non-extant compositional strata remains obscured, "no narrow frith" to cross. But the material processes by which notes from the second-order repositories entered the text have been documented. While he generally wrote out his drafts without using notes, Joyce relied on them for revisions and additions, "sprawled across two beds surrounded by mountains of notes" (SL: 253). Consulting note repositories, he mottled notebooks and notesheets with coloured crayons, and transferred individual entries into his text. In documenting this accretive process, previous studies have focused on either the intertextual origins of individual notes (the 'oxcavations' of Robert Janusko and Gregory Downing, for example, trace the sources of many entries on the "Oxen of the Sun" sheets to books in Joyce's personal library in Trieste) or on the eventual destination of crossed-through notes in the published text of Ulysses (see Philip Herring's annotation to his edition of the notesheets).10 In the wake of the recent acquisitions, however, there are now enough drafts of individual episodes and so many note repositories available that familiar elements of the published text can be seen at the moment of their incorporation into the text of Ulysses and their corresponding notebook sources can be identified. Notes that are worked into a draft occasionally drop out again or become altered in subsequent drafts; this aspect of the compositional process is elided when one can only correlate notebooks with the published text. The presence of "ousted" "possibilites" (U 2.50–51) among the pages of the notebooks alerts one to the other material that did not make it into Ulysses – the great number of uncrossed entries in the notebooks. This so-called paralipomenal material had a curious afterlife after Ulysses was completed; it was transcribed by Mme Raphael in the '30s into what is now Buffalo MS VI.C.7 pp. [136–269].11

The four Ulysses notebooks at the National Library of Ireland were, perhaps accidentally, among some forty notebooks transcribed by Mme Raphael, and it was long thought that they had been destroyed after transcription. Accordingly, in his catalogue of Joyce's manuscripts and letters at Buffalo, Peter Spielberg presented a reconstruction of the beginning and end of this presumed-lost manuscript (he called the original manuscript MS 'VI.D.4' in his catalogue).12 What Spielberg presumed to have been a single document is, in fact, the four notebooks. By comparing Mme Raphael's transcript, Buffalo MS VI.C.7, with Joyce's original four Ulysses notebooks the order of transcription is readily determined.

There seems to be no deliberate order to the sequence of notebooks; presumably they were handed to Mme Raphael as part of a group of notebooks to be transcribed. Few of the entries transcribed from the Ulysses notebooks into Buffalo MS VI.C.7 were used for Finnegans Wake, but the revelation that MS VI.D.4 is in fact four notebooks may have implications for Rose and O'Hanlon's Lost Notebook. The Lost Notebook, Rose and O'Hanlon's reconstruction of a notebook from which Mme Raphael transcribed unused material (into pp. [232–274] of Buffalo MS VI.C.16), has retained the catalogue reference Buffalo MS VI.D.7 after the short reconstruction of it given by Spielberg.13 From their masterly reconstruction, Rose and O'Hanlon show that the original Lost Notebook would have contained elements drawn from Joyce's reading in the Zentralbibliothek while he was living in Zürich in 1918 (and so it is a companion notebook to Buffalo MS VIII.A.5), and from his reading of the London Times, a biographical dictionary, the TLS and dictionaries of slang, among other texts.

Like VIII.A.5, the Lost Notebook draws on Joyce's reading in classical mythology around the subject of the Odyssey: library sources present in both notebooks include Victor Bérard's two-volume Les Phéniciens et l'Odyssée and a French translation of Aristotle's Rhetoric. The library call numbers for both of these are found in the Lost Notebook. Only the Lost Notebook, however, contains notes from Walter Leaf's Troy: A Study in Homeric Geography, which, Rose and O'Hanlon write, first lead Joyce to Bérard (on the other hand, notes from W. H. Roscher's Ausführliches Lexikon der griechischen und römischen Mythologie are present in only VIII.A.5). These sources were undoubtedly consulted in the Zentralbibliothek. But taken together with the London Times and TLS, the sources present in the Lost Notebook constitute rather a mixed bag. While Joyce's eclectic reading habits are well known, the fact remains that the reconstructed notebook divides into two neat portions. The first twenty-three pages relate to the London Times and TLS, and the 'Thomas Fitzgerald' entry from the Dictionary of National Biography; the remainder to the commentaries of Bérard, Leaf and others, and to Ware's text on slang. The latter material points to a notebook compiled, like VIII.A.5, in the Zentralbibliothek. Despite the presence of Zentralbibliothek call numbers on the 'front cover verso', the former material, the first twenty-three pages of the reconstruction, may constitute a separate lost notebook.14

If among the notebooks Mme Raphael transcribed into Buffalo MS VI.C.16 there were thus two Ulysses notebooks, unused portions of the first of these became VI.C.16 pp. [232–247], and unused portions of the second became VI.C.16 pp. [238–274].15 In his preface to the Archive volume containing the facsimile of VI.C.16 Danis Rose defends the D-series divisions (with the exception of VI.D.3 which "cannot with certainty be considered a single workbook").16 The unearthing of the four Ulysses notebooks at the National Library of Ireland has put paid to VI.D.4's unity; if the Lost Notebook is also several lost notebooks Joyce's reading of datable newspapers cannot be viewed as coterminous with his work on classical sources in the Zentralbibliothek Zürich. This speculation does not invalidate Rose and O'Hanlon's findings in The Lost Notebook. Their reconstruction, while it has not received the attention it deserves, deftly corrects Mme Raphael's mistranscriptions and restores hypothetical deleted entries (which she would have skipped over). The Lost Notebook is one of the few critical works which engages with the early phase of the writing of Ulysses, a period about which little is known and studies of which are, regrettably, largely ignored. The elucubrations of Rose and O'Hanlon have dispelled the darkness surrounding the origins of the Lost Notebook's contents; what remains shrouded is the precise nature of this missing document.

As well as being mined for VI.C.16, across three of the four Ulysses notebooks in the National Library of Ireland there is further evidence of large-scale transferral. Almost all of the "Circe" notes in the three later notebooks, NLI MS 36639/4, /5A and /5B, that are crossed in green recur on notesheets BL "Circe" 16 and 17 (BL Add. MS 49975/fol. 20). While a few green-crossed notebook entries do not appear on the sheet,17 (either an oversight on Joyce's part or evidence of another green-cross harvesting stage) almost all of the notes on BL "Circe" 16 can be traced to sources in the notebooks. Likewise for the centre column and some of the margin of BL "Circe" 17. That the notebook entries were transferred from the notebooks onto the two sides of the sheet and not vice versa is evident from Joyce's colour usage: the notesheet entries are crossed in a variety of colours, indicative of transferral during a range of writing phases. Their source entries in the notebooks are, as said above, all crossed in green.18 There is an additional thematic unity to the notes on BL "Circe" 16 and 17 – in the notebooks the entries are grouped under "Circe" headings, those that were transferred onto the sheet are all for, as Joyce called it, the "Messianic scene" (LI: 171). These notesheets thus represent a deliberate recast of notes from, presumably, already heavily-mined sources.

The recto of the sheet as it appears in the British Library Additional Manuscript (BL "Circe" 16) was also the recto from Joyce's perspective – it contains entries copied from NLI MS 36639/5A, and the early "Circe" pages of NLI MS 36639/5B, whereas some later entries from /5B are found on the top of the verso (BL "Circe" 17). The verso also contains entries lifted from NLI MS 36639/4 and a note repository that is no longer extant. Given the thematic unity of the sheet, and the fact that uncrossed entries remained in the notebooks by the time Mme Raphael came to transcribe them, the transfer of material to a sheet was not the result of mining the notebooks for unused material. The recast was the result of selective runs through the three National Library notebooks and some other missing note repository.

By collation of the notebooks with the sheet, both Joyce's methods of culling source notebooks and layering entries onto a recipient sheet are revealed. The following is BL "Circe" 16. The text has been replaced with a lorem ipsum; compare Herring's transcript of the notesheet in Joyce's Ulysses Notesheets in the British Museum for the original text.19 To aid that comparison, the punctuation of the original has been retained and initials in the original (such as LB and IHS) and replaced with capitalized words in the lorem ipsum. The text is colour coded to indicate source notebooks. Entries transferred from NLI notebook MS 36639/5A appear in blue; those from NLI notebook MS 36639/5B are in red. The notebook page from which each note is lifted is given after the entry. Entries, or portions of entries, in black have no known notebook source. As in Herring's transcript, deleted entries are marked between < > and the colour of cancellation is indicated by means of the capital letters R [red]; S [slate]; and G [green].

Some clear patterns of transferral emerge. The earliest stratum of writing on BL "Circe" 16 is, predictably enough, the left column of entries (as usual the left margin, 16:1–20, is later). The left column entries are taken from NLI MS 36639/5A. The entries at 16:22, 24–28 come from the first "Circe"-entitled page of the notebook, p. [10v]. Joyce works through the notebook methodically, from earliest "Circe" page to last, and entries generally remain clustered together on the sheet. For some pages of the notebook, p. [20v] for example, the notes are repeated in an order almost identical to that of the source notebook: the inscription of 16:47–55 follows seriatim the exact order of the notebook.20 For other pages of the source notebook, p. [19v] for example, there is little resemblance between the order of inscription in the notebook and their reproduction on the sheet. To labour the point, among the many other entries on p. [19v], both used and unused, the green-crossed entries appear in the following order: in the left margin of the page, the note that is now at 16:35 appears before 16:34; in the centre block of notes, the notes originally occur in the order BL "Circe" 16:42, 35, 44, 43, 40, 38, 39, 36. While the notes occupy a localized corner of the sheet, the order of their transferral could just as well indicate multiple runs through the source page (crossing and transferring a few notes at a time) as a single run, during which all useful entries were committed to memory and then transcribed in whatever order they came to Joyce.

Either inferred explanation is plausible; more interesting, perhaps, is what happens to the notes during transferral. When Joyce wrote to Weaver, on 20 July 1919, defending the method of "Sirens" (and that of "Cyclops" and "Circe") he said that "[t]he elements needed will only fuse after a prolonged existence together" (LI: 128). In the transferral of elements from notebook to notesheet it is possible to see that fusion in action. The attentive reader may have already spotted that the note at BL "Circe" 16:35 is sourced twice to p. [19v] of /5A in the layout of notesheet 16 given above. That is because 16:35, "R <this stinking goat of Mendes>", comes from two discrete entries on that page of the notebook. Similarly 16:63, "R <ear covers eye, contorts face>", is a composite, the first half of which comes from p. [20v], the second from p. [11v] of /5A (the double sourcing of BL "Circe" 16:71 to /5A p. [27r] is the result of a correction: "G <Bloom 1st Bloom 2nd Sibylline>" in Herring is two discrete entries "Bloom the first, bloom 2nd" and "Sybilline" [sic] in /5A). Other notes on the sheet, 16:28 right column, "R <urine, spittle, saliva, sweat>", and 16:42, "R <you beast>", for example, are truncations of their sources in the notebook. BL "Circe" 16: 44 "R <embroidered prayerbook>" appears as the lengthier green-crossed "hausfrau buried with key (had) embroidered prayerbook" on p. [20v]. Longer entries such as these may help the search for the sources of the notes.

As the sheet fills up, and when he comes to use NLI MS 36639/5B, Joyce fits notes in wherever he can, squeezing into BL "Circe" 16:45, 58 and 59 entries from p. [2r] of the notebook, before moving to the margin for p. [3v]. Interesting features include the duplications of "Hornblower" at 16:6 and 19 and "S. Michael" at 16:3 and 17. The notes at 16:9, "G <row in gallery for & against LB>", 16:12, "G <eagles & palms>", 16:15 and 16, "G <crowded wall falls>" and "R <bone thrown at LB>", are all translations of the Italian entries in Lucia's hand from p. [3v] of NLI MS 36639/5B. Herring's transcript of 16:20, "G <clap clap hands Poldy come home>", might seem at first to be an embellishment of the note "clap clap hands" present in the notebook. What Herring has transcribed as a single note is, however, two notesheet entries – they even occupy two separate lines in the margin. Perhaps the green strikethroughs of both caused him to read them as a single unit. With the body and margin and thus the entire recto saturated, Joyce continues on the verso.  The centre column entries, BL "Circe" 17:33–45, account for the remainder of the notes taken from NLI MS 36639/5B, with the final entry, BL "Circe" 17:44–45, "R <lechery with dissolute grandam, gave example of excessive sensuality>", combining entries on two pages of "Circe" notes in the notebook, pp. [11r–11v]. Interestingly, the first half of the entry, "lechery with dissolute grandam", comes from the second of the two pages; the second half "Gave example of excessive sensuality" from the earlier.

At BL "Circe" 17:46, "R <LB visited medically>", the use of notebook NLI MS 36639/4 begins. 17:81 gives an example of currente calamo development as Joyce begins to reproduce the source note verbatim (from p. [7v] of the notebook) only to alter it: "G <LB funny word  Recorder what is was that?> What is the man saying?". This same page of the notebook, p. [7v], also contains the "Kidney of Bloom" litany (U 15.1941ff.), crossed in green but which does not appear on the notesheet. BL "Circe" 16:77, "G <whores sing litany>", is the only mention of the litany on the sheet and this note is itself a development of the notebook source which mentions only a hymn sung by the whores.21 The earliest next known occurrence of the litany is in Buffalo MS V.A.20.22 Conversely, neither the last few entries in centre column of "Circe" notesheet 17, nor the entire left margin, can be found in any of the extant notebooks.23 This seems to indicate at least one other notebook, one which has not survived. 17:23, the uncrossed "pillar pin of collar", looks like the correction of a note incorrectly transcribed from this missing notebook. Such a notebook or notebooks may have continued onto another notesheet, also missing, for the Messianic scene to which the other Ulysses notebook from the late stage of composition, Buffalo MS V.A.2, may have contributed. Phillip Herring has written:

Most of the [V.A.20] draft originated in "Circe" notesheets 16 and 17, but some seventeen phrases can be found in notebook V.A.2, pages 11 and 12, which means that before item V.A.20 was finished, Joyce had about exhausted his notesheet material and was already mining the notebook. Many more phrases from notebook V.A.2 were added to this section of "Circe" in typescripts and proofs.24

Of the seventeen phrases, five are crossed in green. While Herring speculates, reasonably enough, that Joyce exhausted "Circe" notesheets 16 and 17 before moving onto notebook V.A.2, his use of the notebooks and sheets may have been more interrelated than Herring realized. V.A.2, like the three National Library of Ireland notebooks discussed above, may have been used as the source for another Messianic scene notesheet. Further collation of the preparatory materials is necessary and it still may not show conclusively whether notebook V.A.2 was used as a source for notesheets or after them.

Within the three notebooks that I have shown contributed to BL "Circe" 16 and 17, entries on "Circe"-labelled pages that are crossed in different colours appear interspersed with the green-crossed entries that did end up on the sheet. The former enter the text of "Circe" at every stage: some are even present in the body of the earliest draft, Buffalo MS V.A.19. Joyce was gathering notes for the Messianic scene, not written until summer 1921, during the earliest phase of his work on "Circe". Further work on this process, and this stage of composition, will provide more information on Joyce's move into, what Michael Groden calls, the 'last stage' of the writing of Ulysses.25

In Ulysses in Progress Groden instances two of the three stages of writing with a test case – the early stage with "Aeolus", the middle stage with "Cyclops". Material limitations as well as thematic concerns govern his choice of episodes. The composition of "Aeolus", which "serves as a microcosm of the entire book", comprises an equivalent three stages: the period from May-September 1918 during which the episode was drafted, typed, augmented and dispatched for serial publication; a middle stage of further additions to the typescript for Darantiere for book publication; and a last stage of placard and page proof additions (which included the headlines so familiar from the published text).26 "Cyclops" is not only the episode in which Joyce abandoned the interior monologue, but from then on more care was taken of manuscripts and notebooks and a more complete genetic dossier is extant for the later episodes (certainly this was the case when Groden was writing Ulysses in Progress in the '70s before the appearance of the National Library Joyce Papers).27 For the discussion of the last stage, covering the eighteen months from Joyce's arrival in Paris in 1920 to book publication in 1922, Groden does not single out any one episode. The diffuse treatment he provides stems from the mixture of new writing and revision that characterized Joyce's work on Ulysses after 1920. Since the acquisitions of recent years there are four drafts of "Circe" and six notebooks known to be extant. Now the textual gestation of this episode and Joyce's attendant compositional difficulties as he entered the last stage can be more fully described.

In conclusion, the uncovered material builds on our understanding of Joyce as the "notesnatcher" (FW 125.21–22). The preponderance of note repositories may lead to a 'malestimation' of his role in the production of Ulysses, reducing him to a copyist of his own text. "Scylla and Charybdis" resounds with predictions of its future composition, of a "national epic ... yet to be written" (U 9.309), and includes Stephen's self-exhortations to remember the slight passed on him: "See this. Remember." (U 9.294). Without "slips from the library counter" (U 3.407) he cannot graphically record the incident, but his claim that "in the future, the sister of the past, I may see myself as I sit here now but by reflection from that which I shall be" (U 9.383-85) links him with Joyce creating the scene out of, presumably, such notes in 1918. Stephen's ability to look forward to looking back, he tells us, depends on the simultaneity of selves past, present and future created by "the intense instant of imagination when the mind, Shelley says, is a fading coal" (U 9.381–82). The allusion to Shelley appears not only in Ulysses but also in Portrait and in both versions of the James Clarence Mangan lecture Shelley describes the mind in creation as

a fading coal, which some invisible influence, like an inconstant wind, awakens to intransitory brightness; this power arises from within, like the colour of a flower which fades and changes as it is developed, and the conscious portions of our nature are unprophetic either of its approach or its departure. Could this influence be durable in its original purity and force, it is impossible to predict the greatness of the results; but when composition begins, inspiration is already on the decline. [...] The toil and the delay recommended by critics, can be justly interpreted to mean no more than a careful observation of the inspired moments, and an artificial connection of the spaces between their suggestions, by the intertexture of conventional expressions; a necessity only imposed by the limitedness of the poetical faculty itself; for Milton conceived the 'Paradise Lost' as a whole before he executed it in portions. [...] And let this be the answer to those who would allege the fifty-six various readings of the first line of the Orland Furioso. Compositions so produced are to poetry what mosaic is to painting.28

Mosaic was the correspondence Frank Budgen found for Joyce's writing among the pictorial arts.29 And in 1921 Valery Larbaud compared the preparatory material to "the little boxes of coloured cubes of the mosaic worker".30 Even Joyce himself referred to his marked proofs as "mosaics" (LI: 172). While the occurrence of the mosaic image in 'A Defence of Poetry' may be nothing more than serendipitous, as a description of composition it can only aggravate any sense one may have of Joyce as a mere artisan. Taking Shelley's declaration that "when composition begins, inspiration is already on the decline" to its conclusion, the contents of the National Library notebooks become lifeless repetitions of earlier "significant ... scribbling" that were recorded on more ephemeral material during "the intense instant of imagination". This makes Joyce the amanuensis of his own text. But, by tracing the succession of iterations of a scribbling through second-order and subsequent repositories, one realizes that the recurrence of an entry across preparatory strata is in fact a kinetic process of further creation, a fusing and elaborative process, rather than the labours of a copyist. This Joycean 'rigmarole' of following a thread of transmission, either back to intertextual source texts or forward through subsequent note repositories or deployment in drafts, illuminates Joyce's "dark materials".31

2007 Afterword.

Since this article first appeared in Genetic Joyce Studies further work has been carried out on the Joyce Papers 2002. As we have come to understand the notebooks more, the interrelationships between them and the note repositories previously known to be extant have emerged. Rather than revise the above, I have produced this Afterword to showcase some recent findings. Many more of the entries on the verso of BL Add. MS 49975/fol. 20 (Herring's BL "Circe" 17), for example, have been traced to the National Library notebooks. Their source is the "Eventuali"-labelled pages of the notebooks. To demonstrate this aspect of the transferral a lorem-ipsumed BL "Circe" 17 is reproduced below; again compare Herring's transcript of the notesheet in Joyce's Ulysses Notesheets in the British Museum for the original text.32 Entries transferred from NLI notebook MS 36639/5A appear in blue; those from NLI notebook MS 36639/5B are in red. Those from NLI notebook MS 36639/4 are in green. Entries which come from "Eventuali" pages are additionally boxed in yellow. In this Afterword I also present the current state of research on the dates of the notebooks' compilation and present an errata list for Herring's edition of BL "Circe" 16–17.