Long Lost Manuscript: The True Antecedents Of Professor Lidenbrock, His Nephew Axel And Their Glorious Adventure Underground

William Butcher

Published in The Modern Language Review, Vol. 93, Part 4, October 1998, pp. 961-71

The excitement generated by the recent discovery of a manuscript of Voyage au centre de la Terre is easy to understand. Verne's tale of underground exploration (1864, 1867)(1) is a major literary work and one of his four great novels. It was the first book he wrote under the supervision of his lifelong publisher Hetzel, and the first to appear under the series title Voyages extraordinaires. I am privileged therefore to have been allowed exclusive access to a sizeable proportion of the manuscript; and I record my thanks here for the generosity of the Heritage Book Shop and the new owner of the manuscript.

Until now, the inception of Voyage au centre de la Terre (VCT) has remained totally unknown. The novel seems to be mentioned in only one of the contemporary letters. (2) All Verne's successful novels were 'pre-published' in serial form, generally with many differences from the book edition, except for VCT and Cinq semaines en ballon (1863). Extant manuscripts can be freely consulted for every major work except these two; we can indeed consult two distinct handwritten versions for Vingt mille lieues sous les mers (1869) and two for Le Tour du monde en quatre-vingts jours (1872). (3) These four documents contain important revelations that have changed our conceptions of the books. In the latter case, for example, the opening page, analysed only in 1995, reveals such surprises as anti-Semitism, a probable fourteen-year backdating, an explicit sexual remark revealing Phileas Fogg's real drive, and references to 'Hanovre' and 'le duc de Wellington'. (4) What is more, the earliest surviving fragment mentions clubs, Britain, and 'Fog', in that order. In other words, social observations are paramount when the archetypal modern hero makes his first faltering steps, and no journey, or circumnavigation, or time-limit has begun to be dreamed of. These are fundamental differences.

In its lack of textual information, VCT was thought to be almost unique amongst Verne's novels, until it was reported in Extraordinary Voyages, 1.4 (1994), p. 5 that an earlier version did still exist. Christie's of New York had sold a manuscript of VCT for $266,500 on 7 October 1994; their sales catalogue reports that Verne apparently gave the document to a neighbour, who passed it down to his grandson, and that it remained in private hands until 1991. (5) In March 1996, the Heritage Book Shop of Los Angeles again sold this highly collectable piece to another private individual.

Literary manuscripts have a special appeal, for they reveal the author at his most vulnerable, at the moment of creation on the blank page. Even the struggle to decipher the handwriting and doodles contributes to the feeling of approaching the heart of the mystery of creation. Above all, manuscripts provide information before editors begin interfering. With luck, they may elucidate questions of personality or of structure that have been argued about for generations or even show the moment when Nemo's submarine vessel springs, as it were, from the womb, small but perfectly formed. Of course, the material that survives the editorial process represents what is thought most valuable. But by whom? The studies on this subject (6) have emphasised how deceptive in practice published works usually are. Especially in previous centuries, editors often misread the manuscript and in any case altered it to a greater or lesser extent. Even when authors corrected proofs or revised published editions, they rarely seem to have cross-checked against the final manuscript version and so errors often remained undetected. What does not survive can consequently be just as interesting; and the further back you go, the more you discover. The earlier drafts generally state things less subtly but more clearly; the editor (or author) often removes the material that posterity -- if given the chance -- finds the most creative. The most exciting manuscripts are therefore the original editors' worst nightmares. The older, deeper, and messier the palimpsests, the better.

In Verne's case, although a number of manuscripts have been published and studied, two features have virtually monopolized discussion. A great deal of attention has been devoted to the posthumous novels, stemming from Piero Gondolo della Riva's discovery that Verne's son, Michel, wrote substantive parts of them. (7) The radical amendments imposed by Hetzel to make the endings happy ones have also been pointed out in the case of several non-posthumous novels. (8) Opinions are divided as to whether the editor's forceful suggestions were beneficial, with it sometimes being pointed out that, without his initial 'discovery' of Verne, the Voyages extraordinaires might not have existed at all, but also that his intervention often amounted to censorship of anything that might be considered remotely controversial. (9) But we know surprisingly little about the final stages of the editorial process. Only two sets of proofs seem to have survived: one of them is of the posthumous 'Édom', and the other has not been studied to date. In other words, we know little about the details of Hetzel's editorial corrections, nor, above all, about Verne's reactions to them.

What we do know is that Verne himself extensively reworked his novels. In the 1860s and 70s, at least four or five handwritten versions probably succeeded each other: after a full planning of the plot, a point list was in some cases constructed, covering whole chapters. (10) Then a pencil version was written, presumably amended to some degree, then copied out in ink. This version, in handwriting that is sloping and difficult to read, was normally followed by a mise au net in legible, upright copperplate. Hetzel's son, Jules Hetzel, claimed that five or six distinct sets of corrected proofs were then necessary before the author was satisfied, but it is not clear when this practice dated from. He also commented that the early Verne rewrote more energetically than he did in the 1890s and 1900s. Finally, the pre-publication in serial form and even the first, non-illustrated edition, were also revised: often merely in terms of stylistic detail, but sometimes, especially if the serial audience was an adult one, deleting the sexual, irreverent, or other comments from Hetzel's book editions.

In these terms, VCT is a very good manuscript. Each sheet is a morass of deletions and insertions. Even a glance at the seventy folios that have been made available to me (fos 1-32, 99-104, 131-59, and 191-3) leads to the conclusion that they portray a different Lidenbrock and Axel, a different runic manuscript, and a different underground cavern. Accordingly, a complete publication of the manuscript, with both a reproduction and a transcription, would be a vital addition to knowledge of Verne. While hoping that such an enterprise will soon be possible, my aim here is simply that of summarizing as much significant information about the manuscript as possible.

I shall assume that the reader is acquainted with the 1867 edition, and briefly describe the presentation of the manuscript, and then study two key sections. The first is the opening seven chapters describing Axel, Lidenbrock, and their domestic situation, and recounting their successive attempts to decipher the mysterious parchment that falls out of the ancient book. The second is the five chapters presenting the exploration of the underground cavern and the question of what creatures are to be found there.

Examination of the folios themselves leads to some immediate observations. Verne's writing is clear and regular. Crossing out, however, is done with close-spaced diagonal hatching (), often rendering the text illegible. Each folio has two equal columns, with the text and Verne's corrections appearing on alternate sides of successive folios.

The folios are numbered in sequence from '1' to '193' by Verne himself, and there are chapter numbers from 'I' to 'XLIII'. (11) On folio 101 the chapter heading 'XX' is crossed out; the heading 'XXII' is added to folio 99, and 'XXIII' to folio 103, with both headings corresponding to the 1864 and 1867 editions. Symbols like a stretched square bracket appear in the text at intervals of about four folios, often in the middle of words, apparently indicating a new typesetter, whose name generally appears in the margin. Heritage and Christie's report that although the text itself is in dark brown ink, there are 'comments in red pencil' 'throughout'; but this seems to be a misinterpretation of the unusually expressive names of the typesetters ('Puissant', 'Marteau', and so on).

The large amount of amended text is sometimes misleading, for there are virtually no cases where an insert mark is used to add material above the line. Instead, when the amended text is longer, the whole phrase or sentence up to the end of the line is usually deleted and the new version written out again in the facing column. Such modifications seem to be essentially stylistic, with few changes of great interest. The overwhelming majority of significant textual features fall consequently into three categories: material that has been added, whether in the second column or only in the published edition; material deleted at either of these stages; and text in the manuscript version which differs from the published edition.

As an indication of a typical page, I reproduce here the first folio of the manuscript. The first column represents a transcription, with the 1867 version in the second column.


Le 13 avril 17 xx 22(?) mai 1863, un vendredi lundi mon 24 mai, un dimanche samedi 12 mai 1863, un dimanche, mon oncle, le professeur Lidenbrock, revint précipitamment vers sa petite maison située au numéro 19 Firlander-strasse Königstrasse, l'une des plus anciennes rues du vieux quartier de Hambourg.

La bonne Marthe dut se croire fort en retard, car le dîner au lieu de fumer sur la table, xxx commençait à peine à chanter sur le fourneau xxxxx xxxxxx xxxxxx dans la cuisine de la cuisine.

'Bon, me dis-je, s'il a faim, mon oncle, qui est le plus impatient des hommes, va pousser des cris de détresse.

-- Déjà M. Lidenbrock, s'écria la bonne Marthe stupéfaite, en entrebâillant la porte de la salle à manger.

-- Oui, Marthe, mais le dîner a le droit de ne point être cuit, car il n'est pas deux heures. La demie vient à peine de sonner à Saint-Michel.

-- Mais alors pourquoi M. Lidenbrock revient-il?

-- Il nous le dira vraisemblablement.

-- Le voila! Je me sauve, monsieur Axel, vous lui ferez entendre raison.'

Et la bonne Marthe regagna son laboratoire culinaire. qui xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx appliquxx chaque jour appliquxx xxxxxxxxx vaissxxx xxxxxxxxxxxxxx ayant xxxxxxxxxxxxxxxxxxx restaurant xxxxxx.

Je restai seul. Mais de faire entendre raison au plus irascible des professeurs, c'est ce que mon caractère un peu indécis ne me permettait pas. dont je xxxxxxxxxxxxxxais pas. Aussi je me préparais donc à regagner prudemment ma petite chambre d'en du haut, quand


Le 24 mai 1863, un dimanche, mon oncle, le professeur Lidenbrock, revint précipitamment vers sa petite maison située au numéro 19 de Königstrasse, l'une des plus anciennes rues du vieux quartier de Hambourg.

La bonne Marthe dut se croire fort en retard, car le dîner commençait à peine à chanter sur le fourneau de la cuisine.

'Bon, me dis-je, s'il a faim, mon oncle, qui est le plus impatient des hommes, va pousser des cris de détresse.

-- Déjà M. Lidenbrock! S'écria la bonne Marthe stupéfaite, en entrebâillant la porte de la salle à manger.

-- Oui, Marthe; mais le dîner a le droit de ne point être cuit, car il n'est pas deux heures. La demie vient à peine de sonner à Saint-Michel.

-- Alors pourquoi M. Lidenbrock rentre-t-il?

-- Il nous le dira vraisemblablement.

-- Le voila! Je me sauve, monsieur Axel, vous lui ferez entendre raison.'

Et la bonne Marthe regagna son laboratoire culinaire.

Je restai seul. Mais de faire entendre raison au plus irascible des professeurs, c'est ce que mon caractère un peu indécis ne me permettait pas. Aussi je me préparais à regagner prudemment ma petite chambre du haut, quand

Note especially that the 'Königstrasse' of 1867 is initially 'Firlander-strasse', a further link to the pretty Virlandaise of Axel's dreams (on fo 11 'la jolie firlandaise'). Many other changes are visible within the first chapter. There are stylistic variants from the published text such as 'il avait eu le temps de jeter' (2), (12) the reference to the reader 'voyez-vous' (4), and the interesting fact that the professor's window-panes are, mathematically, 'losangés' (4). If 'Sainte-Claire-Deville [sic]', Verne's friend and author of one source of the novel, is absent from the list of famous scientists to have visited Lidenbrock, an unidentified 'Banral(?)' (3) appears instead.

The portrait 'du professeur de Lidenbrok [sic]' (3) is more ferocious in the opening pages of the manuscript. The first time he is presented, crossings-out frustratingly hide three and a half lines beginning with 'il n'avait pas fait de mal à' (2) and possibly ending with 'agrément' (2). But we can read Axel's damning indictment of his uncle: 'et pourvu qu'il fût instruit, il se moquait bien d'instruire' (2) (replacing the more flattering '[. . .] de paraître instruit' (2)). The irony of Axel's comment that there are one or two professors like this in Germany is reinforced in the manuscript by the sting in the tail 'et même ailleurs' (2). The scurrilous 'Il demeurait avec la bonne Marthe et moi' (4) is typical of Verne's sly humour, and even opens up avenues that Marthe might be Axel's mother! Lidenbrock's financially-challenged Traité de Cristallographie Transcendante is for the moment modestly 'in-octavo' and not grandiosely 'grand-folio' as it will be in 1867. Even his house is sardonically, aggressively, 'au milieu du plus horrible quartier d'Hambourg que l'incendie de 1842 a malheureusement respecté' (4).

The assassination process continues on the following pages. Lidenbrock is sarcastically called 'monsieur', he swears ('Malédiction'), and, to interrupt Axel's daydream, he strikes his shoulder (1867: 'la table'). A hint of homosexuality seems to appear in the description of 'M. Christensen, consul à Hambourg et ami particulier du professeur' (32. 1867: 'Christiensen'). In the published version the professor's locking up of the starving Axel and Martha is perhaps just absent-mindedness, but in the manuscript the intention is made quite clear: 'Quelle clef? La clef de la porte. Non!' (20). Axel's comment is: 'Ma foi, je fus tenté de rire. Mon oncle xxx une sagesse dans ce que j'avais fait la veille! à propos d'une xxx clef' (20). Even Lidenbrock's reaction to the cracking of the code is different: 'Son système nerveux, semblable à celui de la torpille, tressaillait d'électricité. J'aurais eu peur de le toucher. § Ce fut ainsi pendant quelques minutes' (21).

The initial information revealed by the manuscript is therefore striking. Fogg and Nemo were also much sterner figures in the manuscripts. In all three cases, we may detect similarities with Verne's own attitude to his mentors: to his spiritual father, the publisher Hetzel who discovered him but heavily criticised his manuscripts; and to his biological father, whom he considered too strict. The attack on the older generation in the more passionate manuscripts clearly conveys Verne's continuing resentment. The runic document is also different in the manuscript. The Heimskringla in which the parchment is discovered is itself 'in-octavo' (5) rather than 'in-quarto'; and a large number of details in the cryptic message are also significantly different. (13)

But the most interesting revelation is to do with Axel's love for Graüben. The Virland girl is here 'blanche' (11) rather than 'blonde', with a total of eleven lines about her crossed out, including 'chaque jour elle venait avec des fleurs' (11). It is Graüben, rather than Marthe, who emerges after the professor's enraged exit -- and who therefore, after the maid's return to the kitchen, stays in the house alone with Axel, for the one and only time. What they do for sixteen lines (14) (perhaps the same as during the 'quelques détours' (29) on the way home) seems unfortunately to be undecipherable.

It will be remembered that Lidenbrock tests his theory on codes by making Axel write a secret message and then encode it. The original version of Axel's coded message is again illegible. (14) The uncoded form, however, can be made out, and saved for posterity. It is strikingly different from the published version, and calculated to shock Lidenbrock even more: 'Aime moi, ma jolie Graüben, aime moi' (13). Proceeding to physical acts, in other words, is not precluded. Axel had already pointed out that 'moi, j'étais assez léger' (11). Hetzel and Verne senior would most certainly have concurred!

The consequences of Axel's imprudence can now be studied: ' "Qu'est-ce donc que cette Graüben xx," demanda mon oncle pour la seconde fois' (13). Axel's weak reply is: 'C'est un devoir (?) du Johannæum' (13), interestingly implying that he is one of Lidenbrock's students. After the professor returns to the document, there follow three illegible lines, ending with a mysterious and harsh '[. . .] sèchement son interminable soliloque' (13).

Other details are just as illuminating. Axel tells Marthe that they are going a long way down, 'En enfer!' (30); the expedition is called 'la plus insensée [. . .] du XIXe siècle' (7); as they leave the house, Axel holds Graüben in his arms 'sous les yeux un peu attendris de mon oncle' (31)! Altona is described as a 'ville frontière du Danemark' (32), and it is later remarked that 'nous entrions sur le territoire danois du Holstein' (32). Both references to Denmark will disappear in the published editions, the reason being that they are no longer true. Denmark and Prussia fought a war in 1864, resulting in Schleswig-Holstein being occupied and eventually incorporated into Prussia. Undoubtedly because of Verne's last-minute deletion, it has never before been noticed that to publish a book about two Germans visiting Denmark and one of its colonies during a war between the two countries is quite remarkable!

In the other folios, we continue to discover a different Voyage au centre de la Terre. Like many of his contemporaries, Verne's spelling is often non-standard. (15) Many of the dates are also different from those of the published versions. (16) One general category of variants consists of slips, whether corrected in the manuscript or later. Thus the number of active volcanoes on the globe is here amended from 'quinze environ' to 'trois cents environ' (24); the temperature in the heart of the Earth is calculated to be 'deux millions de degrés' (25) (1867: 'deux cent mille'); the underground sea is described as a mere 'lac' (133); the world on the surface, of which the travellers are no longer part, is called the 'globe' (133) (1867: 'la terre' (133)); there is an inconsistent 'brise du sud-est' (133).

Some of the comments do not survive either. Even if their deletion may perhaps streamline the plot, and many are implicit in the rest of the text, we can perhaps regret most of them. Examples include the quasi-uterine idea that in comparison with the immense underground cavern, Guacharo (17) and Mammoth caves are 'de simples grottes, d'étroites cavités!' (132) or the uncharacteristic notion that superman Hans may have snatched a few hours' sleep on the raft. Again, Axel's spontaneous reaction to the notion that the Earth was a vast hollow sphere inside which two heavenly bodies, Pluto and Proserpina, traced their mysterious orbits, is the biting 'Pure imagination! d'un anglais! [sic]' (132), whereas in 1867 he supports it, 'Aurait-il dit vrai?' Finally, a pseudo-learned footnote on fo 142 reads: 'Le professeur Lidenbrock emploie ici la classification proposée pour les poissons fossiles par le professeur Agassiz qui se fonde sur la disposition, la nature et la forme des écailles.' This important comment provides the modern researcher with the previously unsuspected source (18) for the scene of discovery of the first living creatures underground.

Other deletions remove poetic touches reflecting the simultaneously romantic and realist inspiration of Verne's writing. Examples include: the characterization of Axel's dreams as being about 'cosmiques' (132) phenomena, the photographic metaphor of the giant mushrooms as 'cette forêt négative' (133), or the presence underground of 'chutes de hauteur moyenne' (133), all features that contribute to the creation of 'cette nature nouvelle' (136) (1867: 'ces nouvelles régions'). The manuscript initially presented the waters of the Mer Lidenbrock as 'froides mais douces' (134), with an amendment to 'salées' (134) also omitted in 1867; the prehistoric bones are explained as being conserved by the presence of 'carbonate de chaux' (135); the dialogue about whether living monsters might still survive is much longer, concluding with a rude remark about prehistoric monsters: 'Qui nous affirme que leur race ne s'est pas longtemps reproduite!' (135). A final variant is the description of the megatherium as 'fouill(ant) la terre avec ses membres antérieurs' (143), a subtly obscene remark calculated to enrage Hetzel.

The bulk of the variants with the 1867 edition, however, are to do with the professor. The Lidenbrock of the manuscript continues to be generally less sympathetic. At one stage Axel suggests: 'Ah! cette fois, mon oncle, vous vous moquez de votre neveu!' (138) and, just before collapsing from thirst and exhaustion in the bowels of the earth, he notes: 'je le voyais secouer la tête d'impatience' (101), an idea echoed during the underground navigation: ' "Enfin," s'écria mon oncle qui commençait à redevenir le plus impatient des hommes' (140).

The exception to Lidenbrock's unpleasantness is when he is concerned about Axel's first stroll in the underground cave after his illness. For the first and last time, he calls him 'mon neveu' (135); and Axel actually notes: 'Mon oncle me considérait avec attention' (133). In the manuscript it is the professor who points out that the faithful Hans-bach has also reached the cavern. Lidenbrock cedes the reins to his nephew to the extent of allowing him to lead the way into the giant mushroom forest (133) -- with the result that they almost get lost there for ever.

But Lidenbrock reverts to type whenever his anxiety or impatience return. On one occasion he laughs 'nerveusement' (137). He often seems ridiculous through intemperate predictions. Thus he argues that it is 'pas probable' (135) that living creatures survive in the cavern; he emphasises the care with which he has calculated the route (for example, 136), implying that he always knows where he is; he estimates the size of the ocean as 'dix ou quinze lieues tout au plus' (137) (1867: 'Trente ou quarante lieues'), causing him to predict: 'Nous devrons en rencontrer promptement les rivages opposés' (137). Such is his optimism that he also falsely prophesies that they will reach the ultimate goal: 'je suis certain de trouver de nouvelles issues qui nous conduiront à notre but' (136). Even his scientific reasoning may be dubious when he is discussing underground atmospheric pressure: 'Quarante Six atmosphères, et tu vois que cette pression ne nous gêne pas. C'est que nos corps s'y sont habitués peu à peu insensiblement' (136): hence the solution in the published version of simply deleting these sentences.

But the professor does deliver two important monologues in the manuscript:

Oui, mon garçon, ces couches de lignites proviennent non seulement des forêts dont l'Islande fut autrefois couverte, mais aussi des essences tropicales du nouveau monde: ces bois fossiles arrachés des continents de l'Europe ou de l'Amérique et jetés à la côte d'Islande par le grand courant du Gulf-stream, s'y sont lentement minéralisés sous l'action des eaux de la mer, et un cataclysme toujours inexplicable les a entraînés jusqu'ici (139).

'Quel effet à la surface du globe si les Iles Britanniques allaient s'engouffrer dans un abîme de quarante lieues de profondeur!' § 'Bon! Les Anglais inventeraient des machines pour se tirer de là! Mais il n'est point question de pareil d'un semblable bouleversement. Ce qui nous importe, c'est de savoir ce que nous allons faire, désormais' (137).

In the first one, life underground is implied to be derived from that on the surface -- whereas in the published versions the origin of underground life is not discussed. The sci-fi idea of the second is also quite daring, since it concerns, not some past underground realm, but part of the modern world. It connects VCT both with Les Indes noires (1876), whose original version featured a whole parallel world under Britain, and with Hector Servadac (1877), where part of the globe including the British Isles is detached and travels around planetary space. This conversation is also one of the rare occasions where Lidenbrock admits to doubt about his plans.

One important final feature in the manuscript is the use of tenses. The logbook section on the Lidenbrock Sea is entirely in the present in the published versions -- the first example of a major stylistic experiment that Verne carried out over the next thirty years. It is particularly fascinating to note that parts of the manuscript logbook (for example, folios 141-42, 146, 147-48) were originally narrated in the conventional past tenses.

The logbook section in fact contains many important passages worthy of analysis. However, considerations of space unfortunately do not allow me to develop them here. A future article is planned, in which the use of tenses and these passages will receive the attention that they deserve.

The relatively uniform and regular writing, especially compared with the other manuscripts, implies that Verne copied it out from some earlier version (often in pencil on the same sheets). Given the printer's marks, this document must have been the last manuscript stage, although Verne must have intensively corrected at least one stage of proofs. We are extremely fortunate, in sum, that it is so different from the published versions. Commentators have often pointed out how VCT reflects its own existence. At the end of a study of the manuscript, it emerges that the novel here reflects its own earlier version. Many of the features present in the book, such as the indecipherable messages and the texts-within-the-texts, also apply to the unfinished work -- almost as if Verne guessed that a future scholar might pore painstakingly over his manuscript.

Where do the fascinating strata of amendments or deletions made during or after the manuscript stage leave future editions of VCT? The ideal of the modern editor is to publish what the author 'really' intended. But even where changes were made by Verne himself, it was often under pressure from the publisher and even the public. One simpler conclusion is that modern editions should at least correct errors of reading: in addition, the significant variants should be available in the notes of a critical edition. It is also arguable that all current editions need replacing, to reflect Verne's own ideas. But clearly, even if a decision is made to start from the published text, many difficult choices would need to be taken concerning particular passages, often based on subjective appreciations as to the attractiveness of variants.

In particular, many of the deletions and alterations do seem regrettable. An important unanswered question is whether Hetzel saw this manuscript before it was finally submitted: if so, some of the changes in Verne's handwriting might not be entirely his doing, and the need to re-establish the earlier text would be greater. As a result of this uncertainty, it is also difficult to say to what extent modern French and English editions, up to now based exclusively on the 1867 edition, need to be altered. But given the important of this novel, the OUP critical edition needs, in sum, to be completed by a full French or bilingual publication of the manuscript. It is to be hoped that the new owner will permit such an endeavour, which will greatly add to our knowledge of Verne in his vital early years. There are surely monstrous discoveries lurking in those unexplored depths.

At the end of this analysis, then, the element of amazement persists. The most important document appertaining to one of the masterpieces of probably the most-translated author of all time has suddenly surfaced. It is hard to imagine a comparable find happening again. Knowledge of VCT, previously limited to the 1864 and 1867 editions, has suddenly become multilayered and shifting. It is as if the geological and temporal depths depicted in the novel have opened up before our surprised feet. One can legitimately feel a little giddy. But many questions still remain. Where was the precious document until 1991? What lies hidden under the annoyingly close-spaced cross-hatching? Might special photographic techniques enhance the slight difference between the text and the hatching? Is there any hope of exploring possible pencil erasings? And what lies in the other 123 folios?

These are difficult questions, but one fact is certain: Lidenbrock will never be the same again.

Hong Kong Institute of Education William Butcher

1. The first edition (18mo) appeared on 25 November 1864, and was unillustrated. As was usual, the illustrations were not added until the first large-octavo edition (1867). However, the 1867 edition was radically different from the 1864 one, for most of chapters 37-39 were added. These chapters first describe contemporary discoveries in prehistory, and then show them in exaggerated form, in the shape of a herd of giant mastodons, complete with a giant shepherd. (Full information about the new chapters is available on pages 177-87 and 227-28 in my Oxford University Press translation and critical edition of Journey to the Centre of the Earth (1992);  the 3rd edn (1997) contains additional material about the manuscript (pp. Xxxi-xxxii, 231-32). For ease of reference the 1867 edition is used in this article, although the 1864 edition seems to be identical in all respects other than those mentioned above.

2. Although the book is surprisingly not mentioned in the publishing contract of 1 January 1864, the final manuscript must have been finished by 12 August 1864, since a letter to Hetzel of that date enquires when the volume is to go to press.

3. Most of the manuscripts, including MS1 of Le Tour du monde, passed from Jules Verne to his son Michel, to his son Jean Jules-Verne, and then in 1981 to the town of Nantes, with microfilm copies eventually deposited in the French National Library (BN) for public consultation. Le Tour du monde MS2 is also kept in the BN, as are both manuscripts of Vingt mille lieues. Extensive information about the four manuscripts is provided in my translations and critical editions of Around the World (Oxford, Oxford University Press, 1995, pp. xii, xxxii-vi, and 219-47) and Twenty Thousand Leagues under the Seas (OUP, 1998, pp. xxvi-xxxviii and 385-445).

4. Respectively: 'son admission fut [assez] long-temps [discutée] mais enfin il passa sur la recommandation de ses xxxxxx xx, avide, de rotschild et fils chez lesquels il avait un comp crédit ouvert. De là, une certaine [surface] pour ce Fogg'; 'le Reform club, Pall mall, 1858 [or 183x?] xxxxxxxx Burgh'; 'Face n'est pas seul organe expressif'; 'cette institution des arts et des sciences réunies qui est placée sous le patronage direct de la reine et dont le duc de Wellington fut le vice xxxxxx. Il n'appartenait à aucun des nombreuses sociétés de toutes sortes qui pullulent -- [aucun], depuis la société xxxxxxxxx Hanover [équin xsous]'. Where only part of a quotation from Verne's various manuscripts is different from the published editions, the different part is italicized. In all other cases, text quoted in this article is absent from published versions. For ease of reading Verne's cross-hatching is replaced by strike-out. A series of 'x's indicates the approximate number of illegible letters.

5. It was purchased in 1991 by Biblioctopus of Los Angeles, who sold it in 1993 to the Heritage Book Shop of Los Angeles. Heritage sold it to a private collector in Los Angeles, who in turn sold it to Christie's. In November 1994, Heritage bought it back again; and reproduced the first folio of the manuscript in their sales catalogue ('Le 24 mai 1863, un dimanche [. . .]').

6. See, for example Art and Error: Modern Textual Editing, ed. By Ronald Gottesman and Scott Bennett (New York: Modern Language Association of America, 1970) and Scholarly Editing: A Guide to Research, ed. By D. C. Greetham (Bloomington: Indiana University Press, 1995).

7. Principally in Piero Gondolo della Riva, 'A propos des œuvres posthumes de Jules Verne', Europe, 595-96 (nov.-déc. 1978), 73-88.

8. Most notably Capitaine Hatteras, L'Ile mystérieuse and Hector Servadac.

9. Nevertheless, a number of studies have discussed the question of the manuscripts, although usually based on limited readings of them. The most useful are Volker Dehs, 'Les Manuscrits des Voyages extraordinaires: Répertoire des noms disparus', BSJV, 109 (1er tri. 1994), 20-35, Olivier Dumas, 'Défense du style de Jules Verne', and following discussion, in Revue Jules Verne, 3, (1er sem. 1997), pp. 25-38; Bibliothèque de Nantes, Catalogue des manuscrits de Jules Verne (Nantes: Ville de Nantes, 1988).

10. Both these features are visible in the manuscripts of Le Tour du monde.

11. Fol. 99 also seems to read '97' (or '87'?); fol. 100, '98'; and 101, '87'(?); but fol. 102 reads '98' again, fol. 103, '93(?)' and fol. 104, '92'. Fol. 131 replaces '121', and so on up to fol. 159 replacing '149'; fol. 191 similarly amends '181', fol. 192 '182', and fol. 193 '183'. The handwriting is darker and more rounded on folios 99-101, and especially so in the corrections on folios 25-32, 99-100, 103, and 131 onwards. It appears that after the folios up to 102-4 were written, ten or twelve more were added; fos 99-101 must have subsequently been copied out again; and then two new folios must have been added somewhere between folios 32 and 96.

12. Numbers in brackets indicate folio numbers.

13. The text, in Verne's hand, is framed; the gaps between the groups of letters are filled with a runic middle point; the second to fifth lines contain four columns of words; the seventh to nineteenth groups of runes are lightly numbered '7' to '19; and some of the runes are lightly crossed out (7). The transcription in the Latin alphabet (9) shows the same variation in the columns, has all the letters lightly crossed out and dotted underneath and has in the margin a copy of the semi-decoded version for checking (with, however, 'silucoI' rather than 'silucoY'). The published '132 lettres ce qui donne 79 consonnes contre 53 voyelles' is here 'cent trente six deux [. . .] quatre-vingt [sic] [. . .] cinquante deux' (10). The penultimate version of the message has the words 'rota led nec dua atria vir Iira' (13) scrawled in the margin: clearly Verne looking for Latinate words in the jumble of letters. In the event, the manuscript says: 'la dix huitième, la dix neuvieme et la vingtième formaient le mot "dok" ' (15), there are also translations of the words 'ice', 'rota', and 'mutabile' (15), and the comment that 'ice' and 'dok' are easy to connect (16), an allusion to Verne's Voyages et aventures du Captain Hatteras (1864), being written at the same time. The final version has the correct Latin adjective 'audax' (1867: 'audas'), refers (in the French version) to 'Sneffel' rather than 'Sneffels', and adds the word 'Moi,' before 'Arne Saknussemm' (21).

14. In the amended version in the second column (12), the uncoded message contains an erroneous comma between 'petite' and 'Graüben', and therefore the coded message also has one: 'Jmne,b'.

15. Although features such as italics, 'œ' and 'æ', and 'Nedåt' (104) or 'chûtes' (133) are meticulously penned (these two diacritics disappear in print), there are such slips of the pen as 'au contraires' (11). Accents in general are often missing ('fût' is invariably 'fut'); and amongst individual words are mistakes such as 'graphytes', 'sinet', 'hypotèse', 'couremment', 'dégrés', 'quand à', 'Fourrier', and 'acquéreraient'. (Given the difficulty of reading much of the text, correct spelling is used in this article.)

16. Amongst other examples, 'le 28 juin' (30) becomes 'le 22 juin' when published, but 'le 15 13 août' (139) reverts to 'le 15 août' in 1867. The margin of folio 18 reads 'Lundi le 25 mai Mardi le 19 mars', without it being clear whether this is for Verne's own information or for publication.

17. 'Guacharo' Caves is spelled correctly in the manuscript; the published 'Guachara' is an error.

18. Louis Agassiz (1807-73) was a Swiss-American zoologist and geologist and anti-Darwinian author of a five-volume Recherches sur les poissons fossiles (1833-44), four-volume Contributions à l'histoire naturelle des Etats-Unis (1857-62), and a Voyage au Brésil (1868), which inspired Verne's La Jangada (1881).