Chapter I(a)
The Pryftan Fragment

The original page from a student essay upon which Tolkien scribbled down the words ‘In a hole in the ground there lived a hobbit’ does not survive, but a substantial fragment of six pages (three sheets) from the original manuscript has been preserved. This I have dubbed ‘The Pryftan Fragment’, after the name given the dragon at this earliest stage of the story. The fragment lacks both a beginning and an end, but it does form a continuous text which is given below.

It is not clear now how far this initial stage of composition carried the story. According to Tolkien’s later recollections, the story halted before the end of the first chapter and may indeed have stopped at the point where the fragment ends.¹ Nor is it clear what happened to the missing pages. They may have been given to some friend, as Tolkien gave away other bits of Hobbit material – specifically, the original of the Mirkwood picture (Christopher Tolkien, Foreword to the 50th Anniversary Hobbit, p. x; Pictures by Tolkien, plate 37) and a very fine unused picture of Smaug flying around the Lonely Mountain (Foreword, p. xiii). Nor was The Hobbit the only one of his works he treated in this way: he gave an elaborate illuminated manuscript of his still-unpublished poem ‘Doworst’ to his friend R. W. Chambers² and similarly gave away both the manuscript of and copyright to the then-unpublished poem ‘Bilbo’s Last Song’ to his secretary Joy Hill in gratitude for her years of service.³ Inherently unlikely as it may seem from our historical perspective that Tolkien would give away the single most famous page of manuscript he ever produced,⁴ his generosity in other cases on record makes it a distinct possibility.

Or the missing pages may have been deliberately destroyed by Tolkien after being translated into typescript. Contrary to legend, Tolkien did occasionally destroy manuscript material when, as in this case, it was rough draft workings that had definitely been superseded by a later fair copy or typescript. For example, in both The Book of Lost Tales (cf. BLT I.45, 64, 130, 174, 203; BLT II.3, 69, 138, 146, 221) and in sections of the Lord of the Rings material, Tolkien would often draft a passage in pencil, then write a revised form of the text over it in ink, typically afterwards erasing whatever pencilled jottings remained, completely obliterating the initial version. While it may be argued that such extraordinary measures were forced upon him by paper shortages in wartime, no such explanation will suffice in the case of Mr. Bliss. The little hand-made booklet reproduced in facsimile in 1982 is a carefully made fair copy that clearly required extensive preliminary drafting for both the art and the text, yet only a stray leaf or two bearing sketches for some of the illustrations survived to accompany the hand-painted manuscript book when it arrived at Marquette in the late 1950s; it seems clear, in this case at least, that Tolkien himself discarded the missing rough draft material. Furthermore, Christopher Tolkien notes an analogous case of missing rough draft for the 1937 Quenta Silmarillion, where only a small portion of the pages upon which Tolkien worked out the revisions incorporated in this text survive (HME V.199). Then, too, the Hobbit manuscript itself shows one clear, unambiguous case where Tolkien ripped a page of Ms. in half; the piece which survives does so only because its back was re-used for some outline notes.⁵ Tolkien kept a great deal of his own manuscript, probably so he could reconstruct the text should the final version be lost or mislaid⁶ (and of course because this would enable him to re-use elsewhere ideas and elements that had dropped out of this particular story), but even he did not keep everything.

Finally, and most probably, the missing manuscript pages may simply have been lost by accident. According to Tolkien, C. S. Lewis on two separate occasions accidentally destroyed the only copy of a story Tolkien had loaned to him to read (Carpenter, The Inklings [1978], p. 48), and other mishaps doubtless occurred. Perhaps it would be better not to speculate on how the missing pages were lost, but to ask how the surviving pages happened to be preserved. Two of the sheets (four pages of text) from this first stage of composition (Marq. 1/1/22:1–4) came to Marquette in June 1957, mixed in with the rest of the Hobbit manuscript and typescripts but very distinct from them in the style of Tolkien’s handwriting and the type of paper used. The third sheet was retained by Tolkien, either inadvertently or because it bore the first sketch of what came to be known as Thror’s Map.⁷ Reproduced in facsimile in Christopher Tolkien’s Foreword to the 50th anniversary edition of The Hobbit (Unwin Hyman 1987, pp. ii–iii),⁸ it did not join its fellows at Marquette until July 1987 (MSS–1 Tolkien, Mss. 1/1/1).

While the Marquette processors made no record of how the papers were arranged upon arrival, we are unusually fortunate in that some surviving correspondence relating to the sale casts valuable light upon both Tolkien’s own recollections concerning the papers and on how he had them stored before they came to Marquette.⁹ Tolkien initially told Bertram Rota, the London bookseller who acted as Marquette’s agent in the sale of the manuscripts, that there was no actual manuscript, only the ‘original typescript’ sent to the printer, the corrected proofs, and his illustrations for the book (Rota to Ready, 10th January 1957). After ‘looking through his cupboards’ he turned up the original Farmer Giles typescript (‘There is no hand-written version of this work, which was composed on the type-writer’) and asked for ‘a bit longer to dig around and see if he finds any more bits and pieces concerning “The Hobbit” and “The Lord of the Rings”’ (Rota to Ready, 5th May 1957). By 13th May, he had discovered the manuscript of Farmer Giles, the very existence of which he had forgotten so completely as to deny a week before that there ever had been one (Rota to Ready, 13th May 1957); a month later when Rota arrived in Oxford to collect the first installment of the papers for shipment to Marquette, he discovered that ‘Tolkien has found . . . more than we expected . . . When I wrote on May 5th I reported that Tolkien said there was no hand-written manuscript of “The Hobbit”. Now he has found it . . .’ (Rota to Ready, 13th June 1957).

Even allowing for mistakes or misunderstandings on Rota’s part (evidenced elsewhere in his letters to Ready), it is quite clear from this account that Tolkien’s memory of the Hobbit manuscript, superseded as it had been by the typescript some quarter-century before, was understandably vague. It is also clear that the material was not all kept in one file, but scattered among his papers,¹⁰ and that Tolkien had some difficulty in locating and pulling all the pieces together. In fact, as we shall see, some pieces evaded his search and are still retained by the family to this day.¹¹

The following is the complete text of the surviving fragment; comments and observations follow the transcription. I have provided punctuation as necessary and corrected a few obvious slips (e.g., replaced ‘the the’ with simply ‘the’) but otherwise have edited this first draft as lightly as possible.

As they sang the hobbit felt the love of beautiful things made by hands and by cunning and by magic moving through him; [added: A fierce and jealous love, the desire of the hearts of dwarves. Then] something Tookish awoke within [>inside] him, and he wished to go and see the great mountains and the seas, the pine trees and the waterfalls, and explore the caves [of<] and wear a sword instead of a walking stick.^TN1 He looked out of the window. The stars were out in a dark sky^TN2 above the trees. He thought of the jewels of the dwarves shining in dark caves. Then in the wood beyond the Water a flame leapt up – somebody lighting a wood fire probably – and he thought of plundering dragons lighting on his quiet hill and setting it all in flames. Then he shuddered, and quite suddenly he was plain Mr Baggins of Bag-end Under-Hill again.

He got up trembling, he had [added: less than] half a mind to fetch the lamp, and more than half a mind to go out to fetch it^TN3. and hide in the cellar behind the beer-barrel and not come out again till all the dwarves had gone away.

Suddenly he found them all looking at him with eyes shining in the dark.^TN4 ‘Where are you going?’ said Gandalf in a tone that seemed to show he guessed both halves of the hobbit’s mind.^TN5

‘What about a little light?’ said Bilbo.

‘We like the dark’ said all the dwarves: ‘Dark for dark business. There are many hours before dawn’.

‘Oh’ said Bilbo and sat down again in a hurry – he sat on the fender and knocked the poker and the shovel over with a crash.

‘Hush’ said Bladorthin. ‘[Silence in the>] Let Gandalf speak.’

[This is some part of what Gandalf said, > And this is how he began >]

‘Bladorthin, Dwarves, and Mr Baggins.^TN6 We are met together in the house of our friend and fellow-conspirator, this most excellent and audacious Hobbit – praised be his wine, and ale –’ (but this praise was lost on Bilbo Baggins who was wagging his mouth in protest against being a fellow-conspirator and audacious, but no noise would come he was so upsettled). ‘We are met to discuss our plans. [Before we go forth>] We shall start soon before the break of day on our long journey – a journey from which some of us [cancelled: may] (or all of us with the probable exception of Bladorthin) may never return. The object of our journey is [all>] well-known to all of you. To Mr Baggins, and to one or two of the younger dwarves (Kili and Fili at any rate – if I am not mistaken) the exact situation [may be unknown>] at the moment may [be >] require explanation.’

This was Gandalf’s style. In the end he would probably have said all he wanted to, and left a little time over for some of the others to have a word. But on this occasion he was rudely interrupted.

Poor Bilbo could not bear it any longer. At ‘may never return’ he began to feel a shriek coming up inside, and very soon after it burst out like a whistling engine coming out of a tunnel.

All the dwarves sprang up knocking over the table. Bladorthin struck a blue light on the end of his magic staff and [by the >] in its glare they saw the poor little hobbit kneeling on the hearthrug shaking like a jelly (a jelly that is melting). Then he fell flat and kept on calling out ‘struck by lightning, struck by lightning’ over and over again. And that was all they could get out of him for a long while. So they took him and laid him on the drawing room sofa with a lamp [added: and a drink] beside him, and went back to their dark business.

‘Excitable little man’ said Bladorthin as they sat down again. ‘Gets funny queer fits, but one of the best, one of the best – as brave [> fierce] as a dragon in a pinch –’ (if you have ever seen a dragon in a pinch you would realize that this was only poetical exaggeration applied to any hobbit, even the Old Took’s great uncle Bullroarer who <was> so large he could sit on a Shetland pony; and charged the ranks of the goblins of the Gram Hill [> Mount Gram] in the battle of the Green Fields of Fellin^TN7 and knocked their king [> King Fingolfin]’s^TN8 head clean off with a wooden club. It sailed two hundred yards and went down a rabbit hole, and in this way the battle was won [added: by checkmate] and the game [> games] of Golf [added: & chess] invented simultaneously).^TN9

In the meanwhile the dwarves had forgotten about Bullroarer’s gentler descendant, and he was recovering in the drawing room.^TN10

After a while (and a drink) he crept nervously to the door of the parlour. This is what he heard – Dwalin speaking.

‘Humph, will he do it, d’you think. It is all very well for Bladorthin to talk about his hobbit being fierce, but one shriek like that in a moment of excitement when we really get to work [> to close quarters] will [> would] be enough to kill the lot of us. Personally I think there was more fright in it than excitement, and if it hadn’t been for the secret sign on the door, I should have been sure I had come to the wrong house, as soon as [added: I] clapped eyes on the [added: fat] little fellow bobbing on the mat. He looks more like a grocer than a burglar!’

Then Mr Baggins turned the handle & walked in. Took had won. He would [cancelled: <rather>] go without bed and breakfast to be thought fierce, and never be called ‘a fat little fellow bobbing on the mat’ again. Many a time afterwards the Baggins part regretted his decision and his strange behaviour at that moment; but [added: now] he went right and put his foot in it without a doubt.

‘Pardon me’ he said ‘if I have overheard [part >] some words that you were saying. I cannot pretend to understand it all, but I think I am right in believing that you think I am no good. I am not – but I will be. I have no magic signs on my door and I am sure you have come to the wrong house – but treat it as the right one. Tell me what you wish me to do and I will try it – if I have to walk from here to [cancelled: Hindu Kush] the Great Desert of Gobi and fight the Wild Wire worm<s> of the Chinese. I had a great-great-great uncle Bullroarer Took and –’

‘We know we know’ said Gloin (in embarrassment) ‘holed out [added: -checkmated] in one in the battle of the Green Fields. But I assure you the mark was on the door. The mark was here last night. Oin found it and we gathered tonight as soon as we could for the mark was fresh.’

‘I put it there’ said Bladorthin from the darkest corner. ‘With my little stick I put it there. For very good reasons. [cancelled: Now let’s get on] – I chose [cancelled: this] Mr Baggins for the fourteenth man and let anyone say He is the wrong man or his house the wrong house who dares. Then I will have no more to do with your adventure, and you can all go and dig [added: for] turnips or coal.’

‘Bilbo my boy,’ he said turning to the hobbit. ‘Fetch the lamp, and let’s have a little light on this dark matter.’

On the table in the light of a big lamp with a red shade he spread a parchment map. ‘This I had from Fimbulfambi (?)^TN11 – your grandfather, Gandalf,’ he said in answer to the dwarves’ excited questions. ‘It shows the Black Mountain and the surrounding country.^TN12 There it is, that dark blob [> lump > tangle]. Over here is the Wild Wood and far beyond to the North, only the edge of it is on the map, is the Withered Heath where the Great Dragons used to live.’

‘We know all that’ said Balin. ‘This won’t help – there is a picture of a dragon in red on the Mountain, but [that won’t make it any ea[sier] >] it will be easy enough to find him without that.’

‘There is one point’ said the wizard ‘which you haven’t noticed, and that is the secret entrance. You see that runefn">^† on the East side and <the> hand pointing to it from the runes below [cancelled: them]? That marks an old secret entrance to the Mountain’s halls.’

Written at the bottom of this page is the following footnote:

^† Don’t ask what that is. Look at the map, and you will see [added: that] one

This clearly refers to the ‘F’ rune marking the secret door on Fimbulfambi’s map (see Frontispiece).

‘It may have been secret in the old days’ said Gandalf ‘but [how do you >] why should it be any longer. Pryftan has dwelt there long enough to find out all there is to know about those caves by now!’

‘He may – but he can’t have used it for years and years!’

‘Why so [> Why]?’

‘Because it is too small. “Five feet high is the door, and four abreast [> three abreast] may enter it” say the runes. But Pryftan could not creep in a hole that size, not even when he was a young dragon, certainly not in the [days >] after he had devoured so many of the maidens of the valley.’

‘[How >] It seems a <pretty> big hole’ piped up Bilbo. He loved maps, and in the hall there was a large one of the County Round (where he lived), with all his favourite walks marked on it in red ink. [This was quite exciting>] He was so interested he forgot to be shy and keep his mouth shut. ‘How could such an enormous [hole >] door (he was a hobbit, remember) be secret’.

‘Lots of ways’ said Bl. ‘but which one of them we don’t know without looking.^TN13 From what it says on the map I should say that there is a closed door which looks just like the side of the mountain – the ordinary dwarf’s way (I think I am night?)’

‘Quite’ said Gandalf.^TN14 ‘[added: But] This rather alters things. There are fourteen of us – unless you are coming, Bladorthin. I had thought of going up along Running River from the Long Lake – [if ever we could rea[ch] >] if we can get so far! – and so to the Ruins of Dale Town. But we none of us liked the idea of the Front Gate. The River runs out of that great door, and out of it the Dragon comes too. Far too often.’

‘That would have been no good’ said Bl. ‘without a mighty warrior even a hero. I tried to find one but I had to fall back (I beg your pardon, but I am sure you will understand – [cancelled: this] dragon slaying is not I believe your hobby [> speciality]) – to fall back on Mr Baggins [> little Bilbo]’.

‘A [> The] burglar’ said Dwalin. ‘Precisely’ said Blad, not allowing Bilbo time to object.^TN15 ‘I told you last Thursday it would have to be a burglary not a battle, and a burglar I promised to find – I hope no one is going to say I put the sign on the wrong door again.’ He frowned so frightfully at Bilbo that the little man daren’t say anything though he was bursting with questions.

‘Warriors are very busy fighting one another in far lands’ went on Bld. ‘and in this neighbourhood [are >] there are none or few left of men dwarves elves or hobbits^TN16 not to speak of heroes. Swords <in the world> are mostly blunt, and axes used for [> on] trees and shields for dishcovers, and dragons comfortably far off. But burglary is <I think> indicated in any case by the <presence> of the back door.’

‘What is your plan’ then they all said. ‘To go to the back door; sit on the step and think of one – if one does [added: not] sprout up on the way’ said the wizard. ‘There is no time to lose – You must be off before day break and well on your way – Dwarves

In the top margin of this sixth and final page of the fragment, Tolkien wrote the following list of dwarves’ names:

Dwalin Balin Fili Kili Dori Nori Oi^TN17 Oin & Gloin Bifur Bofur Bombur Gandalf

It will be noted that all the dwarves are named here, and in the order of their appearance in the typescript made from the now-vanished opening pages of this chapter, even down to the detail of Fili naming himself before Kili (their names being transposed in the final book; cf. DAA.39). From the ink, this list of names probably dates from the original period of composition or shortly thereafter. Much later, probably at the same time as he added the note to the other side of this sheet that it was the ‘Only page preserved . . .’ (cf. Note I above), Tolkien added the following in pencil at the end of the line:

NB Gandalf was originally

Chief Dwarf (=Thorin) and

Gandalf was called Bladorthin.

Here the fragment ends, in mid-sentence at the bottom of a page, and it is probable that no more was written at this stage. But from what we have we can, after the fashion of Sir Thomas Browne,^TN18 make some deductions about the contents of the missing pages that once preceded it; see the commentary that follows.

TEXT NOTES

1 This sentence originally continued with a semicolon followed by the word ‘and’, but these were cancelled and the period inserted.

2 Originally this was followed by the word ‘and’ and the beginning of another word that either started with h-, tr-, or possibly th-; these were cancelled at once and the sentence continued as shown.

3 Here ‘go out to fetch it’ was replaced by ‘pretend to fetch it’. Earlier in the sentence, in the haste of capturing the thoughts before they got away, Tolkien actually wrote ‘and more and half a mind’, which I have altered editorially to ‘and more than half a mind’.

4 This sentence was revised to read ‘Suddenly he found the [singing >] music & song had stopped and they were all looking at him . . .’

5 Tolkien originally began to write ‘Bilbo’ here – i.e., ‘both halves of B[ilbo’s mind]’.

6 This appositive, which originally followed ‘fellow conspirator’ on the next line in the manuscript, was bracketed and marked for insertion at this point. Tolkien originally began the line with ‘Dwa’ (i.e., Dwarves), which was immediately cancelled; similarly, initially the name ‘Bladorthin’ was followed by an incomplete phrase (‘Bladorthin of the’), but this too was immediately cancelled and we have no way of knowing what the wizard’s completed title or derivation might have been.

7 The name of this battle (in the published book simply ‘the Battle of the Green Fields’) underwent several changes in this earliest manuscript mention. First Tolkien wrote ‘the Battle of the’ followed by a cancelled, illegible word of four or five letters that ended in -ll (possibly ‘Bull-’?). Then he resumed with ‘Green Fields of Fellin’. Later he cancelled ‘Fellin’ and wrote ‘Fao’ above it, but struck this out in turn (probably at once, without completing the word) and replaced it with ‘Merria’. None of these names appear elsewhere in the legendarium, the closest approach being the Merrill, one of the rivers of Rivendell (HME VI.205). I cannot identify the meaning of these names, nor the language(s) to which they belong, although Taum Santoski left behind a linguistic note associating Fellin with Noldorin fela (cave) – cf. Finrod Felagund (‘Finrod, lord of caves’) – and suggesting a connection between Merria and Quenya merka (‘wild’); cf. ‘The Etymologies’, HME V.381 (under the root PHÉLEG-) and 373 (under the root MERÉK-). In any case, it appears not to have been a direct translation of ‘Green Fields’, since the Elvish words for ‘green’ are laeg or calen (Sindarin) [Letters pp. 282 & 382] and laiqa (Quenya) [‘The Etymologies’, HME V.368], respectively, each of which has deep roots to the early days of the mythology.

8 The name ‘King Fingolfin’ is written in the left margin alongside this line. See pp. 15 & 24–5 for commentary on Tolkien’s unexpected use here of this elven name, which in The Silmarillion is given to the High King of the Noldor, one of the greatest of the elf-princes fighting in the wars against Morgoth.

9 This long parenthetical kept expanding as Tolkien wrote; originally he intended it to end after ‘exaggeration’, then after ‘any hobbit’, but deleted the closing parenthesis each time and in the event failed to ever provide one, so I have added it editorially at what seems the appropriate place.

10 This sentence was altered through deletions to read ‘In the meanwhile Bullroarer’s gentler descendant was recovering in the drawing room.’

11 The question mark is in the original, and probably indicates Tolkien’s uncertainty about the appropriateness of the name. Like the other dwarf-names in this chapter, ‘Fimbulfambi’ is Old Norse and comes from the Elder Edda; see pp. 15 & 24 for the name’s source and meaning.

12 Here we have, for the first and only time, the original name for the landmark that plays such a large part in the second half of the book. Tolkien originally wrote ‘the Black mountain’, then capitalized ‘Mountain’ and cancelled ‘Black’ to give just ‘the Mountain’, the designation it thereafter retained within the opening chapter; unnamed on the map, it does not gain its full name as the Lonely Mountain until early in what is now Chapter III (cf. p. 111 for its first appearance in the draft manuscript, and DAA.87 for the corresponding published text).

Just before the word ‘country’ later in the same sentence, Tolkien began to write a word which seems to have started with a capital ‘K’; if so, then this might be the first (abortive) reference to the Kingdom under the Mountain.

13 The rest of the page, from this point on, is the first map of the Mountain: see the Frontispiece and the commentary beginning on p. 17.

14 Here Tolkien originally began to write a name beginning with D, but immediately cancelled it and wrote Gandalf instead. While this might have been either Dwalin or Dori, the former is more likely, since the old dwarf had already taken part in the conversation and would do so again a few paragraphs later.

15 Tolkien originally began the next sentence

‘Yes’ said the

then changed this to

‘It w[ould]

before finally settling on

‘I told you last Thursday it would have to be a burglary not a battle . . .

16 Tolkien struck a line through part of this sentence: ‘. . . and there are none or few left of men dwarves elves and hobbits not to speak of heroes’. Presumably the cancellation of the word ‘left’ was inadvertent, and he intended the revised line to read ‘and there are none or few left, not to speak of heroes’.

17 Christopher Tolkien reads this name as Oi rather than Ori, the name we would have expected, and notes (Foreword, page iv; personal correspondence, CT to JDR, 4th November 1994) that Ái is a dwarf-name appearing in the Völuspá, one of the component poems that make up The Elder Edda. See the commentary on the dwarves’ names in Appendix III for more on this and other variants.

18 ‘What song the Sirens sang, or what name Achilles assumed among women, though puzzling questions, are not beyond all conjecture.’ – Sir Thomas Browne, Urn Burial [1658].

(i)
The Lost Opening

In general structure, the lost opening must have paralleled that of subsequent versions fairly closely, however much it may have differed in detail. We know from other accounts that the opening line was either exactly the same as the familiar one in the published text or some close variation on it: e.g., ‘In a hole in the ground lived a hobbit’ or even ‘At the edge of his hole stood the hobbit’ (see ‘The Chronology of Composition’, pp. xii – xiii). References in the fragment to Bilbo’s ‘Tookish’ side show that the Took/Baggins dichotomy was already well-established, even at this early stage, and the motif of the ‘Unexpected Party’ is clearly present. The two references to Bilbo as the fourteenth member of the party make it quite clear that Bladorthin’s withdrawal from active participation at some point had been foreseen from the outset and was not a later development (although, as we shall see, the exact timing of his departure remained undecided for a considerable time). The dwarves’ personalities are, for the most part, much as they remain in later drafts, though it is interesting to note that more of them participate in the discussion than will later be the case. Thus Dwalin, Gloin, Balin, and Gandalf all have speaking parts in rapid succession, and references to Fili and Kili’s youth and Oin’s having been the one to find the secret mark on Bilbo’s door bring more of the full cast into play; Tolkien seems to be trying to make use of the full ensemble of his characters. Later streamlining will reduce the number of dwarven speakers in this passage from four to two, reassigning Dwalin’s speech to Gloin and Balin’s to Gandalf, retaining the reference to Fili and Kili while dropping all mention of Oin’s contribution. While some interesting detail is thus lost, Tolkien’s decision to focus the active roles on only a few of the dwarves (primarily Gandalf, Balin, Fili, and Kili, with lesser roles delegated to Dori and Bombur) makes it much easier for someone listening to the story to keep the characters straight. We might regret that some of the dwarves are relegated to such obscurity that they have virtually no speaking parts at all,¹ but overall the story is strengthened by the simplification.

At least one poem, the dwarves’ song about their lost treasure, was already part of the story, as may be deduced from the opening line of the fragment. A single line of this song (‘To claim our long forgotten gold’) survives by chance, thanks to Tolkien’s thrifty re-use of paper: he originally wrote this line on the first surviving sheet of the fragment (Marq. 1/1/22:2), then crossed it out, turned the page upside down and over, and used its reversed back (1/1/22:1) to draft the next bit of text (the section immediately following the now-lost poem; i.e., the beginning section of our fragment).

(ii)
Nomenclature in the Pryftan Fragment

The most startling thing about the fragment, from the point of view of readers familiar with the later published text, are the unfamiliar names given to several of the major characters and places: Pryftan instead of Smaug, the Black Mountain and Wild Wood instead of the Lonely Mountain and Mirkwood, Bladorthin instead of Gandalf, and especially Gandalf the dwarf instead of Thorin Oakenshield (son of Thrain son of Thror). Tolkien prided himself on his nomenclature (radio interview with Denys Gueroult, BBC, 1965; see also JRRT to SU, 16th December 1937; Letters p. 26), and rightly so; it is a point on which he excels any other writer of fantasy, even Dunsany and Morris – he was able to embrace the exoticism of the one and plainstyle of the other as the occasion warrants without ever losing his own distinctive touch. In point of fact, assigning the name ‘Gandalf’ to a dwarf and ‘Bladorthin’ to a wizard is quite appropriate. The dwarf-name comes from the same list in the Elder Edda, the Dvergatal, that provided the names of all but one of the dwarves who accompany Bilbo on this quest;² like them, it is Old Norse. Fimbulfambi, the original name tentatively given to the King under the Mountain, the character who would later become Thror the Old, likewise comes from Old Norse; this time from the bit of eddic lore known as the Hávamál.³ Bladorthin, by contrast, is Elvish⁴ – specifically, Sindarin, or ‘Noldorin’ as it was called at the time (see Note 13 below for the distinction between Gnomish, Noldorin, and Sindarin) – and as such helps distinguish the wizard from his associates, just as the very English-sounding ‘Bilbo Baggins’ sets the hobbit apart from the rest of the company.⁵

No less surprising is the use of the name Fingolfin for the goblin-king killed by Bullroarer Took: the first of many borrowings that explicitly link Mr Baggins’s world to that of the mythology. While the name was undoubtedly appropriate in form, containing as it does the key ‘golf’ element necessary for the joke, it nonetheless comes as a great shock to readers familiar with the great elven-king as he appears in The Silmarillion, the ‘Sketch of the Mythology’, and ‘The Lay of Leithian’ to have it assigned, even briefly, to a goblin-king.⁶

It seems quite clear that Tolkien is here, as elsewhere in The Hobbit, drawing names from already-written tales and fragmentary sketches with little concern for how well their new use corresponds to that of their first appearance. This is quite understandable when we remember that these were, after all, unpublished and mostly unfinished stories known to (at most) two or three other people. We know from other evidence that Tolkien spent a great amount of time crafting names for his characters (in the Lord of the Rings papers, an entire page of rough workings survives to show how Tolkien worked his way through over thirty rejected names for his ranger Trotter (i.e., ‘Strider’) before eventually coming up with Aragorn). Any artist might want to find a way to reuse unpublished material arrived at with such effort, and Tolkien was thriftier than most; the totality of his work also has a unity unusual in any author. His mythology filled his mind to the extent that it is no surprise to find him borrowing names, ideas, and themes from it in a new work; indeed, it would be surprising if he did not. As he himself said in 1950, ‘though shelved . . . the Silmarillion and all that has refused to be suppressed. It has bubbled up, infiltrated, and probably spoiled everything (that even remotely approached “Faery”) which I have tried to write since. It was kept out of Farmer Giles with an effort, but stopped the continuation. Its shadow was deep on the later parts of The Hobbit . . .’ (JRRT to SU, 24th February 1950; Letters p. 136).

Several other miscellaneous points of the fragment deserve commentary. The golf joke was redoubled by later additions so that the goblin king’s death provided the occasion for the creation of not one but two new games for survivors of the battle: golf and chess. Fortunately, Tolkien soon thought better of this rather forced jollity and it vanishes without a trace at the next stage, where the original joke was restored to its full glory. References to ‘the Water’ and Bilbo’s map of ‘the County Round’ (not, note, ‘The Shire’ – the latter conception did not yet exist) show that the essential neighborhood surrounding Bag-end (already so named) is much as it remains. Indeed, for all the small but significant differences, it is surprising how closely the final story follows this first hasty draft, sometimes even in phrasing. One interesting detail that did not survive is contained in Bladorthin’s cancelled line about his efforts to find a hero or warrior to join the expedition, only to discover that the warriors are all ‘busy fighting one another in far lands’ – echoes of the wars of Beleriand in the Silmarillion tradition, perhaps? – while as for heroes ‘in this neighbourhood . . . there are none or few left, of men, dwarves, elves, or hobbits’. The idea of heroic dragon-slaying hobbit warriors is an intriguing one, and may have influenced both the elusive figure in the Lord of the Rings manuscripts of Peregrin Boffin, or Trotter, the hobbit ranger who eventually metamorphosed into Strider (cf. HME VI.371 & 385), as well as Tolkien’s original plan for the climax of The Hobbit, described in Plot Notes B & C, that it would be Bilbo himself who would slay the dragon (see pages 364 & 496).

(iii)
The Geography of the Tale & The First Map

One of the most remarkable things about this fragmentary draft, and one of the ways in which it most differs from the published text, is the casual use of place-names taken from the real world: China, the Gobi Desert, Hindu Kush, even the Shetland Islands (one assumes, from the mention of the ponies). At first, this gives the reader the impression that Mr Baggins’ world is a totally different place from the legendary world of The Silmarillion. But this impression is deceptive, especially when we consider that in the early stages of the mythology Luthany, the lonely isle later known as Tol Eressëa, was England itself (BLT I.24–5); Kortirion among the trees the city of Warwick; Tavrobel the village in Staffordshire where the Tolkiens lived in the early days of their marriage. As Tolkien originally conceived it, his stories told the mythic history of England and the neighboring lands; a conception he never completely abandoned.⁷ Christopher Tolkien warns us time and again in his edition of The Book of Lost Tales that just because an element drops out of the later versions of one of his father’s stories does not necessarily mean that the conception had been abandoned; often it simply shifted into the background, held in abeyance. The same is undoubtedly true of this element of The Hobbit.

That Bilbo’s world, the lands of The Silmarillion, and our own world are all one (albeit at different points in history) is demonstrable through many of Tolkien’s explicit statements:

‘Middle-earth’, by the way, is not a name of a never-never land without relation to the world we live in (like the Mercury of Eddison⁸). It is just a use of Middle English middel-erde (or erthe), altered from Old English Middangeard: the name for the inhabited lands of Men ‘between the seas’. And though I have not attempted to relate the shape of the mountains and land-masses to what geologists may say or surmise about the nearer past, imaginatively this ‘history’ is supposed to take place in a period of the actual Old World of this planet.

—JRRT to Houghton Mifflin Co., 30th June 1955; Letters p. 220.

Those days, the Third Age of Middle-earth, are now long past, and the shape of all lands has been changed; but the regions in which Hobbits then lived were doubtless the same as those in which they still linger: the North-West of the Old World, east of the Sea.

—LotR. 14; italics mine.

The Lord of the Rings . . . takes place in the Northern hemisphere of this earth: miles are miles, days are days, and weather is weather.

—JRRT to Forrest J. Ackerman, June 1958; Letters p. 272.

Thus a real constellation like the Big Dipper (or, as Tolkien preferred to call it, the Sickle), set in the sky by Elbereth ‘as a challenge to Melkor . . . and sign of doom’ (Silm.48) appears on Fimbulfambi’s map and can be seen by Frodo in the night sky over Bree (LotR.191); the calendars in Appendix D of The Lord of the Rings are calculated to fit a planet with exactly Earth’s orbit, and so forth. It is dangerous to extrapolate backwards from The Lord of the Rings into The Hobbit, but it seems safe to conclude that Bilbo’s story shares this one characteristic at least with the works that both preceded and follow it: all are assumed to take place in the legendary past of our planet. The ‘legendary’ part is worth stressing, since Tolkien was writing fantasy, not pseudo-history or pseudoscience à la Ignatius Donnelly or Immanuel Velikovsky. This liberates him from any obligation to make the details of his setting consistent with ‘what geologists may say or surmise’ and to replace real prehistory (insofar as we know it) with a feigned private history of his own devising.⁹ Like the Britain of Geoffrey of Monmouth and Aegidius of Ham, Bilbo’s world is full of anachronisms, from policemen on bicycles to mantle clocks; in this The Hobbit resembles works like Dunsany’s ‘The Bird of the Difficult Eye’ and ‘The Long Porter’s Tale’ (both in The Last Book of Wonder [1916]) more than, say, the neo-medieval romances of William Morris.

If Bilbo’s impassioned ‘Tookish’ speech makes it clear that his world is firmly identified with our own, can it likewise be tied to the imaginative geography of Tolkien’s earlier tales? The answer, I believe, can be found by turning to Fimbulfambi’s Map. Although differing in significant details from the final version, it is remarkable how many permanent elements were already present and persisted from this first hasty sketch, which shows the mountain laid out two-dimensionally like a starfish. Among these details are the location of the Front Gate (labelled ‘FG’ on the map), the secret door (marked with an ‘F’ rune, as promised in Tolkien’s footnote on the preceding Ms. page; see p. 9), the ‘Ruins of Dale Town’, and something of the surrounding countryside: the River Running (which originally had an eastward course), the ‘WILD WOOD’, and the ‘WITHERED HEATH’.

The Mountain’s north-east spur was only separated by a brief gap from another height that disappears off the map to the northeast, probably a chain of mountains – a feature that soon vanished from the Lonely Mountain maps (cf. ‘Thror’s Map I’, Plate I [top]) but remains in both the earliest sketch Wilderland map (part of Plot Notes B; see pp. 366–7) and also in the more polished Wilderland Map that accompanied the ‘Home Manuscript’ (Plate I [bottom]), which brings the Iron Hills down to almost connect with the Lonely Mountain.

Later the original easterly course of the River Running was scratched out and the river is instead made to bend south once it passes the ruins of Dale Town. Several other new features are added as well: Lake Town upon the Long Lake (the former labeled on the map but not mentioned in the text, the latter named in the text but not labeled on the map), Mirkwood (originally along the bottom or southern border of the map, later expanded up the left margin to form the western border and then the whole southwest corner of the map), and the marshes between them. The Forest River, complete with northern bend before it empties into the lake, is present but not named. The addition of all these extra features makes this first map the ancestor not just of Thror’s Map but of the larger-scale Wilderland map as well. Finally, a third stage of additions to this map, probably made when the story had reached what is now Chapter XI (i.e., about two years after this first drafting), pencils in the dwarves’ first camp just to the west of the mountain’s southernmost spur (the height that would later be called Ravenhill). At the same time, Tolkien added the side view of the mountain (also in pencil) in the lower right-hand corner of this page; compare it with the more careful, nuanced version directly based upon it, drawn to accompany the ‘Home Manuscript’, which is reproduced on Plate II (top).

Mirkwood and the Wild Wood are probably simply two names for the same place: the great primeval forest that once covered most of Europe, one of the remnants of which bears the name the Dark Forest to this day. As Tolkien notes in a letter to his eldest grandson,

Mirkwood is not an invention of mine, but a very ancient name, weighted with legendary associations. It was probably the Primitive Germanic name for the great mountainous forest regions that anciently formed a barrier to the south of the lands of Germanic expansion. In some traditions it became used especially of the boundary between Goths and Huns . . .

—JRRT to Michael George Tolkien, 29th July 1966; Letters p. 369.

However, this is not just a borrowing from historical scholarship, as in the case of the dwarf-names (although it is that as well), but also from Tolkien’s literary roots: William Morris, perhaps his chief role model as an author, and one of the few whose influence he was proud to acknowledge,¹⁰ used the name Mirkwood in his novel The House of the Wolfings [1888] for the name of the great forest where the Germanic woodsmen who are the heroes of the story won a battle against the invading Romans. Furthermore, Carpenter tells us that this book was one of those Tolkien bought with the prize money he received when he won the Skeat Prize for English in the spring of 1914 (Carpenter, p. 69), just at the time when he was creating the first poems of his mythology.

Can Mirkwood or the Wild Wood be tied to any of the great forests in Tolkien’s early mythology? Certainly Beleriand itself was originally called ‘Broseliand’ (later emended to ‘Broceliand’) in ‘The Lay of Leithian’ (HME III.160), the 1930 Quenta (HME IV, pages 107–8, 115, 122, 125, and 131), and on the first Silmarillion map (ibid., between pages 220 & 221); a name clearly borrowed from the great Forest of Broceliand of Arthurian legend.¹¹ A much better candidate, however, is Taur-na-Fuin (also known as Taur Fuin or simply Taurfuin), the Forest of Night. Comparison of the first Silmarillion map in Volume IV of The History of Middle-earth with Fimbulfambi’s Map shows a striking parallelism in the former’s placement of Taur-na-Fuin and Dor-na-Fauglith, the ruined plain to the north between Beleriand and Thangorodrim also known as Anfauglith, and the latter’s Wild Wood and Withered Heath; if the two maps were blended, the Mountain would probably be to the southeast of the highlands later know as Dorthonion, just off the eastern edge of the map, near where Tolkien would later place the Hill of Himring (cf. the published Silmarillion map). We are told by Bladorthin that the Withered Heath is ‘where the Great Dragons used to live’, and I think it more than coincidence that Anfauglith is where Glorund, Ancalagon the Black, and all the rest of Morgoth’s dragons are first seen by the outside world.

This parallelism is strengthened by the figure of the Necromancer. In ‘The Lay of Leithian’ we are told that, after his defeat by Luthien and Huan, Thû the necromancer took the shape of a vampire (that is, a vampire bat) and flew

to Taur-na-Fuin, a new throne

and darker stronghold there to build.

—‘The Lay of Leithian’, lines 2821–2822; HME III.255.

The Quenta (circa 1930) simply states laconically that ‘Thû flew in bat’s form to Taur-na-Fuin’ and that after the destruction of his tower and Felagund’s burial there ‘Thû came there no more’ (HME IV.iii). In the published Silmarillion [1977] this becomes ‘Sauron [= Thû] . . . took the form of a vampire, great as a dark cloud across the moon, and he fled, dripping blood from his throat upon the trees, and came to Taur-nu-Fuin, and dwelt there, filling it with horror’ (Silm.175). As we shall see (p. 73), a cancelled manuscript reference early in the Second Phase makes explicit that the Necromancer whose tower Beren and Lúthien destroyed and the Necromancer in whose dungeons Bladorthin encountered Gandalf’s father are one and the same. Hence the conclusion seems inescapable that Taur-nu-Fuin, the forest to which Thû the necromancer fled to build ‘a new throne and darker stronghold’ and Mirkwood, where the Necromancer defeated by Beren and Lúthien now dwells at the time of Mr Baggins’ story, are one and the same. Its geographical location shifts as the ‘Third Age’ of Middle-earth slowly takes shape in its own right through the writing of The Hobbit itself, eventually (as the second layer of changes to Fimbulfambi’s Map show) developing its own landscape that could no longer be fitted easily into the older geography, so that ‘Mirkwood’ comes to occupy a central position in Wilderland (which now seems quite distinct from Beleriand) closer to that of the Forest of Doriath on the old Silmarillion maps rather than Dorthonion (the place of which is eventually taken by the Grey Mountains on the later Wilderland maps).

A final piece of evidence for the original identification between the Mirkwood, Taur-na-Fuin, and the Wild Wood can be found in the illustrations. The first edition of The Hobbit featured a halftone of Mirkwood (see Plate VII [top]) that was unfortunately dropped from later reprintings. Comparison of this drawing with a painting Tolkien did of Taur-nu-Fuin (H-S#54) to illustrate the story of Túrin the Hapless shows that the two are identical, tree by tree. Only incidental details have changed: the two elves in the painting are not of course in the later drawing, replaced by a large spider and several extra mushrooms. By itself, this could be taken as just another example of Tolkien’s characteristic self-borrowing, but in conjunction with the other evidence, it seems conclusive: the two forests look the same because they are the same; the same patch of woods at two different points in its history.

Two curious points about the map itself should be noted. The first is the compass rose:

Fig. 1: The compass rose from Fimbulfambi’s Map

The pattern on top is clearly meant to represent the Big Dipper (the dark marks to the left of the constellation as reproduced in the Frontispiece are simply stray stains and splotches on the Ms.), and thus indicates north: the shift in orientation to turn the map on its side and place East at the top would not occur until much later. To the South is the sun. East is indicated by the sun rising above some sort of archway or gate, probably the Gates of Morn mentioned in ‘The Tale of the Sun and Moon’, which is described as ‘a great arch . . . all of shining gold and barred with silver gates’ (BLT I.216). West is marked by a three-tiered mountain, possibly meant to suggest the as-yet-unmentioned Misty Mountains (which do indeed lie west of the Lonely Mountain) but more probably the Mountain of the World, Taniquetil, in the Uttermost West. Only some two years earlier Tolkien had painted the magnificent picture of Mount Taniquetil (H-S#52) featured on the front cover of both the Bodleian centenary exhibition catalogue J. R. R. Tolkien: Life and Legend and of Artist & Illustrator, having already appeared in Pictures (as Plate 31). This famous painting shows Taniquetil as a tall peak surrounded by lesser heights which, in profile, would look very like the small icon on the compass rose.¹²

The other puzzling feature about the map is that it does not, in fact, correspond to the one described in the accompanying text. Specifically, Balin points out ‘a picture of a dragon in red on the Mountain’, when there is neither dragon or any trace of red ink on this map. Furthermore, Bladorthin quotes the runic inscription, translating it as ‘Five feet high is the door, and four [> three] abreast may enter it’. In fact, literally transcribed, the runes on the map itself read as follows:

FANG THE

SECRET PASAGE

OF THE DWARVES

The runic system is the same as that followed in the published Hobbit – i.e., Tolkien used the historical Anglo-Saxon runes commonly known as the futhark rather than one of his invented alphabets such as the Cirth. The use of ‘Fang’ here is interesting, because it is an early example of his usage in The Hobbit of his invented languages (specifically, Gnomish, the language that eventually evolved into Sindarin).¹³ It is also an explicit link of the new story back to Tolkien’s earlier legendarium, the tales that were eventually published as The Silmarillion. In the earliest version of the legendarium, The Book of Lost Tales (1917–20), one of the two races of dwarves is known as the Indrafang or ‘Longbeards’; indeed, use of the word ‘fang’ for ‘beard’ persisted into The Lord of the Rings (Fangorn, ‘Tree-beard’). And, as we shall learn in the third chapter, Gandalf and all his companions belong to the Longbeards, or Durin’s Folk as they were later called, a fact first adumbrated by this runic passage.

Below the runes and rather sinister-looking, long-nailed pointing hand was added a version of the text Bladorthin cited, along with the first draft of both the visible message on the map and what became the moon-runes passage:

five feet high is the door and three may walk abreast

Stand by the grey stone when the crow knocks and the rising sun [will >] at the moment of dawn on Durin’s Day will shine upon the keyhole.

This second sentence was at some later point bracketed; the word ‘crow’ was replaced with ‘thrush’ and ‘keyhole’ changed to just ‘key’ (but the cancelled part of the word was underlined, possibly indicating it was to be retained after all). Then the whole sentence was cancelled and replaced with the following:

Stand by the grey stone where the thrush knocks. Then the setting sun on the last light of Durin’s Day will shine on the key hole.

The latter, of course, corresponds more closely to Elrond’s spontaneous translation of what he reads from the map in Chapter III; see p. 116.

Taken together, these discrepancies in a rough draft text would mean little – even after publication, the words on the map and their translation in the text did not agree until this was put right in the second edition (see p. 749) – were it not for the specific reference to something that’s not there; i.e., the image of the dragon in red on the mountain. Given Tolkien’s fondness for ‘handouts’ – actual physical copies of documents seen by his characters, later examples of which include Bilbo’s contract (plate two of the Frontispiece) and the pages from the Book of Mazarbul (a similar impulse can be seen expressed in the Father Christmas Letters) – it’s quite possible that he made a fair copy map that is now lost. Perhaps Tolkien’s choice of words in his comment to Auden that ‘for some years I got no further than the production of Thror’s Map’ (see p. xii; italics mine) suggests a rather more elaborate map than this rough sketch drawn directly into the pages of the ongoing narrative, but this seems too slender a basis upon which to build much. If it ever existed, the lost map must have been quite similar to the next surviving map (Thror’s Map I; see Plate I [top]), which bears the label ‘Thror’s Map. Copied by B. Baggins’, retaining as it does the Northward orientation of Fimbulfambi’s Map and representing the Mountain with a very similar style of hatching. Here the runes translated by Bladorthin are in place, and the back of the map has the moon-runes drafted on the first map. Furthermore, the newer map shows the dragon right on the center of the Mountain, exactly as described by Balin, in contrast with the final published version of the map (Thror’s Map II), where the dragon is flying above the mountain, not resting on it, and the whole scene has been rotated 90 degrees to place East at the top of the map (contrast Plate I with DAA.50 & 97). But for all that, ‘Thror’s Map. Copied by B. Baggins’ cannot be the map Balin and Bladorthin are referring to, since the proper names written on it (Thror and Thrain) did not arise until near the end of the Second Phase, some two years after the Pryftan Fragment was abandoned.