The following material, found by Christopher Tolkien tucked into one of his father’s copies of the second edition of The Hobbit (the sixth printing of 1954), was discovered too late to be incorporated into the original edition of this book. But because of its great interest in demonstrating how vividly Tolkien visualized each scene, because it provides details nowhere else available, and because most of it seems to belong to a ‘Seventh Phase’ and thus represents Tolkien’s last bout of work on the book, I have added it here.
The material in question consists of eighteen pages, written at various times but mostly dating from circa 1965–66, to which I have given the designator Ad.Ms.H.S.xx—that is, Additional Manuscripts Hobbit Supplemental, page xx. The bulk of this material is concerned with the runes used in The Hobbit: Tolkien provides several charts that would enable readers to decipher the various writing systems found on the book’s maps and in the illustration of Smaug on his hoard (‘Conversation with Smaug’; see Plate XI), as well as several more drafts of Bilbo’s Contract in different styles and notes on dwarven modes of writing, punctuation, the representation of numbers in tengwar, and so forth. Only some of this material is included below, since its full presentation and analysis belongs rather to a study of the evolution of Tolkien’s writing systems such as that being carried out by Arden Smith in the ongoing periodic journal Parma Eldalamberon. Also included in these miscellaneous draftings, and presented in full below, are a timeline of events from the opening of the Secret Door to Thorin and Bard’s parley, a detailed description of the terrain immediately outside the Lonely Mountain’s Front Gate, Tolkien’s responses to two queries regarding proposed changes to the text, and a descriptive list of all the dwarves with their characteristic colours. Taken all together, these represent Tolkien’s final sustained work on the book.
This timeline fills the bottom half, top margin, and long left margin of a single page [Ad.Ms.H.S.8]. Immediately following upon Tolkien’s detailed description of the area around the Lonely Mountain’s Front Gate (see Part ii: Notes on a Parley, below), it gives a day-by-day account of the events during the busy fortnight following Thorin & Company’s discovery of the Secret Door.
As before, doubtful words are indicated by french brackets < > while illegible passages are represented by open elipses (. . .).
| Day | Event TN1 |
| 1 | Discovery of door,TN2 day 1. sunset. B’s [added: first] converseTN3 with Smaug [> steals cup] evening/night. Smaug comes out. Company take refuge inside tunnel. S. hunts ponies. |
| 2 | Noon. Bilbo visits Smaug, returns late afternoon. Evening <company> . . . <the> tunnel.TN4 Night S. smashes the door and goes off to Lake Town. |
| 2 | night. Smaug destroys Lake Town, but is slain by Bard. Night spent in tunnel. |
| 3 | Bard sends messages to Elvenking. [cancelled: But he has already <been informed> of D’s death and Bard’s <?men> find him already moving an army towards the L. Mountain <5th> dayTN5] Dwarves explore Smaug’s Hall. |
| 4 | Dwarves come out Front Gate near noon. The[y] proceed towards Ravenhill “5 hrs march” p. 254TN6 They sleep in guardsroom on Ravenhill. They see gathering of birds.TN7 |
| 5 | Bard’s messengers meet the Elvenking already on the move with his host towards the L. Mountain. The EK. turns <?south> towards Lake-town. Dwarves <see> even more birds. Roac son of Carc tells them the news (Smaug’s death 3 nights ago at moon-rise). Roâc [is sent >] is bidden to send messages to Dain. Dwarves return to the Mt. |
| 6 | |
| 7 | Elfhost reaches Lake Town. |
| 8 | ?Fili & Kili sent out to bring back stores. |
| 9 | |
| 10 | |
| 11 | <head> of the <allied> host reaches the ‘rock gates’TN8 where the river flowed into north end of Long Lake. F & K return and report approach of allies. |
| [Note: The entries for the following days are written in the left margin.] | |
| [12] | |
| [13] | “Then came a night”TN9 pm day 13 when the Dwarves saw the lights of the host away <South> [> East] in Dale. |
| 14 | <the allies> move camp <nearer> and crossed old road. Scouts of Men & elves approach Gate . . . <spread> on either side of <M>.TN10 |
| 15 | Bard and Thorin parley. . . .TN11 |
This brief timeline is self-explanatory; in it, Tolkien details the events of the fifteen days between the discovery of the Secret Door and the beginning of the Siege of the Mountain. Had Tolkien continued it, it would no doubt have extended to include the Battle of Five Armies, the death of Thorin, Bilbo’s departure, and perhaps even some highlights from his return journey. Combined with the 1960 Hobbit material given in ‘Timelines and Itinerary’, especially parts ii, iii, and iv, we see how, although in fragmentary form and never finalized, a reasonably complete timeline of Bilbo’s journey could be constructed. For how this timeline relates to those previous efforts, see in particular Text Note 3 on page 828, which if accepted would equate ‘Day 1’ above with 19th October.
Text Notes
TN1. I have provided these headers for the sake of clarity.
TN2. The word ‘door’ here is underlined, and written above it is a D-rune 
, as on Thror’s map.
TN3. The word appears to be ‘converse’ (i.e., ‘Bilbo converse with Smaug’), but may be intended for ‘conversa[tion]’ instead (i.e., ‘Bilbo’s first conversation with Smaug’).
TN4. Of the two illegible words represented here by the ellipses, the first ends in ‘-ts’ and the second might be either ‘open’ or ‘again’.
TN5. This cancelled sentence was recast into the entry for the fifth day; see below. The ‘5’ seems to have been written over another letter, which cannot now be read.
TN6. The reference is to Balin’s comment near the end of Chapter XIII: ‘Not at Home’ in the second edition, where he mentions
‘. . . the old look-out post at the South-West corner of the Mountain.’
‘How far is that?’ asked the hobbit.
‘Five hours march, I should think. It will be rough going . . .’
—second edition, page 254; cf. DAA.299
TN7. The sequence of these last two sentences is reversed, so as to read ‘They see gathering of birds. They sleep in guardsroom on Ravenhill’, which more accurately represents the sequence of events in the story.
TN8. The allusion is to the last line of Chapter XIV: ‘Fire and Water’ in the published book: ‘It was thus that in eleven days from the ruin of the town the head of their host passed the rock-gates at the end of the lake and came into the desolate lands.’ See DAA.313.
TN9. The reference is to a passage in the middle of Chapter XV: ‘The Gathering of Clouds’:
“There came a night when suddenly there were many lights as of fires and torches away south in Dale before them.
“ ‘They have come!’ called Balin. ‘And their camp is very great . . .’ ” (DAA.320).
TN10. The illegible passage here consists of roughly three words, the last of which has been replaced by an equally illegible word. In any case, the allusion is to the human and elven host taking up siege quarters in the valley outside the Front Gate. In the original published book the relevant passage read ‘That day the camp was moved and was brought right between the arms of the Mountain’; in the 1966 Longmans/Unwin edition – the edition with which this ‘Seventh Phase’ is associated – that line was changed to read ‘That day the camp was moved to the east side of the river, right between the arms of the Mountain’ (DAA.320; emphasis mine).
TN11. The second half of this entry, consisting of four or five words, is unfortunately illegible to me. The second word is perhaps days, and the final word seems to begin with st- and ends in an ascender followed by a descender, but I cannot make out the sense of the passage. The relevant section of the text is the account of Thorin’s parley with Bard at the end of Chapter XV: ‘The Gathering of Clouds’, ending in his total rejection of Bard’s terms and the beginning of the Siege; cf. DAA.322–4.
The following account [Ad.Ms.H.S.6–8], to which I have given the name ‘Notes on a Parley’, provides an example of Tolkien’s endless creativity, his ability to expand upon practically any point in his legendarium—in this case, turning his attention to the exact lay-out of the area immediately below the siege-wall Thorin & Company constructed blocking the Front Gate.
At the top of the first of these three pages (numbered ‘1’, ‘2’, ‘3’ by Tolkien in the margin of each page) Tolkien has scrawled
Hobbit
Arrangement at the Lonely Mt. dwarfroads
and <approachments> Time-table of Events
from discovery of secret door
Immediately below this come two lines of runes:
which, when transliterated, read
This school edition is published
by Longmans Green and company
This shows that this material clearly dates from the period (June–September & November–December 1965) when Tolkien was working on the 1966 Longmans School Edition/Unwin Books (trade paperback) edition of The Hobbit, since that edition did indeed include some changes to the descriptions of areas in the vale of Dale; cf. Hammond Descriptive Bibliography page 37; DAA.300, 319, & 320. It seems likely that Tolkien wrote out the account which follows to clarify the lay-out of the area in his own mind, rather than for possible inclusion in the book.
According to my visualization of the scene (which was and remains clear) the parleys between the allies and the dwarves took place on a ledge at about the level [added: and to the right] of the top of the new defence-wall across the gate, where Bilbo and the Dwarves stood. The picture (+214)TN1 was meant to fit that, but the later [added: defence] works were not shown, since <it> was a picture of what the [scouts saw as >] dwarf-scouts saw as they first approached the Front Gate from outside. ((There is <thus> nothing in the picture to show how Bilbo, when evicted and ‘swung down’ reached Gandalf and the other emissaries. I think there was a narrow ledge outside the wall above the culvert made for the issue of the stream; at its right (West) it ran towards the new path upon the bottom ‘ledge’ shown in the picture 214 but left a gap bridge[d] temporarily by a wooden gangway.))TN2
The old road had run on the other left (east) side, where in the picture there is a slope of boulders projecting into the stream. The course for some distance from the Gate had been destroyed, and left a slope of boulders, by the Dragon. It had gone on (and still did so further on) down into the valley along the east-bank of the river, and so on away from the Mountain, eastwards towards the Iron Hills. To reach the old town of Dale (of which now only a few ruins remained) a road branched off westward to a place where at [a] <stony> shallow at the beginning of the great eastward bend of the stream there was [> had been] a bridge. At the western end of the bridge a road went east <to> Dale west towards Ravenhill, and south <to the> river[,] along the bank (left) of which it turned[,] passed under the feet of Ravenhill[,] and went away South again towards Long Lake. ((The bridge should be marked \\ as in my . . . copy))TN3 The “allies” TN4 moved up along both sides of the river (p. 272), but before long [written above this: next day] their camp and main force (272) was made east of the river to command the old outward road—[cancelled: <the other>] approach from the east at any <?other> point needed to cross the river. But the Ford by the ruined bridge was also guarded by Elves (see p. 280).TN5
The Dwarves did not ‘remake’ or reopen the ruined stretch of the old road, because they had not time; and what they for the moment needed was a way out for [cancelled: themselves >] <any> of their own number sent [cancelled: out] on an errand or scouting. They thus cleared and levelled a space at the end of the lower shelf on the west cliff (to their right), which was at about the height of the top of their new wall across the mouth of the Gate. (They expected parleys to take place) From that they had <hurriedly> cleared away some of the obstuctions along the ledge so that a dwarf (or hobbit) on foot could proceed along it. How far? Evidently not all the way to Ravenhill and then down the steps and path from it into the valley:TN6 <For> Bilbo had only 5 hours in all, and in the dark he could not have got to the camp even in that time. There must have been some way down from the ledge into the valley (see p 280),TN7 allowing B. <room> to get back to the stream, and so come to the Ford—which [he] would <now> have to cross <the> <?other> way—eastward to go [> get] to the camp. It is probable that the ledge was <partly> artificial made <in> the old days, where people could walk. And guards be posted <?surveying> the valley: and no doubt between the Gate and Ravenhill there were several ways down from the ledge to the valley (though narrow and easily defended). This fits description of battle in which Elves were evidently posted along the ledge as well as on the tops of the ridge of the ‘Southern Spur’ (eg the one ending in Ravenhill.) but were all to get down to attack enemies in the valley.
Dain naturally approached the Mountain by the old road from the east. It was a weak point in the dwarves’ plan that, for outside help to reach them, Dain would have to force the pass between the river and east cliff, and then cross the river by the Ford and so come to the gate along the narrow ledge, where his 500 would be strung out in a line. But the plan was made by Thorin, while he still hoped that Dain would arrive before the ‘allies’ arrived or had occupied Dale. (He probably also did not suppose that in any case anyone would try to stop Dain, or attack himself. He was very conscious of his rights [> dignity] as lawful ‘K. under the Mountain’, and convinced of the justice of his position and his ‘right’ to deal out his treasure at his own free will and <estimate>.
Note: The expedition to Ravenhill [made bef[ore] >] was made on the day of the Dwarves’ emergence from the F. Gate. <?took>. . . 5 hours.TN8 <it went> by old road.
The text ends here, in the middle of the third page [Ad.Ms.H.S.8], with the Timeline given above crowded onto the rest of this page. In keeping with his preferred role as chronicler of feigned history, it is notable how Tolkien carefully combs through the existing text for clues about the geographical feature he is describing, even citing specific page references, rather than simply re-writing the passage to reflect his new visualization of the scene.
Text Notes
TN1. This is the picture ‘The Front Gate’ (DAA.256; H-S#130), which appeared on page 214 of the second edition Hobbit, near the beginning of Chapter XI.
TN2. These two sentences, which I have enclosed in double parentheses, are bracketed by Tolkien. A final cancelled word, ‘which’, originally followed ‘gangway’.
TN3. This sentence is enclosed in brackets. The single illegible word before ‘copy’ might be ‘control’; in any case, Tolkien is referring to a master copy or check copy of The Hobbit in which he has entered all the changes and corrections he has made to the second edition text.
TN4. We might also note in passing that ‘allies’, the designator he applies several times to the forces opposed to Thorin & Company, is a loaded term for any veteran of the Great War and survivor of the Second World War, wherein the Allies (England, France, the US, etc.) fought the Central Powers and the Axis, respectively.
TN5. This is an allusion to Bilbo’s encountering, and being captured by, elven look-outs when crossing the stream on his self-appointed emissary to Bard and the Elvenking (DAA.328).
TN6. Written in the left margin opposite this line is the notation ‘Sc. 255’. This is a reference to the description of this terrain on page 255 of the second edition (corresponding to DAA.300).
TN7. This is another allusion to the passage already referred to in Text Note 5, specifically, ‘It was very dark, and the road after a while, when he left the newly made path and climbed down towards the lower course of the stream, was strange to him. At last he came to the bend where he had to cross the water . . . The bed of the stream was there shallow but already broad, and fording it in the dark was not easy for the little hobbit. He was nearly across when he missed his footing on a round stone and fell into the cold water with a splash . . .’ (DAA.328; italics mine). For an account of the construction of this new path, see DAA.319.
TN8. I cannot read the cancelled word or phrase preceding ‘5 hours’.
These two brief pieces were apparently written in response to queries on specific points by a typesetter or proofreader at the printer. Both are found on the same sheet [Ms.Ad.H.S.9] and, Christopher Tolkien notes, ‘typed on the back of a page from an engagement calendar for 16–22 August 1964’.
For the first query, the passage in question comes from the last sentence of the fourth paragraph of Chapter V: ‘Riddles in the Dark’. Groping around for matches, Bilbo’s hand ‘came on the hilt of his little sword – the little dagger that he got from the trolls, and that he had quite forgotten; nor do the goblins seem to have noticed it, as he wore it inside his breeches.’ (1st ed., page 81; italics mine). Christopher Tolkien also notes that, in the copy of The Hobbit in which these notes were discovered, the relevant passage is changed to ‘nor fortunately had the goblins noticed it’, the reading which appears in the Unwin Books edition (page 64) and subsequent editions (cf. DAA.116). The following reponse shows Tolkien’s rationale for the change, and reveals how carefully considered his revisions could be.
Queries, p. 64 (former [page] 81)
[added in pencil: p. 80 in 3rd ed.]
(a) seem. where the estimate of probability is made by the narrator with reference to a past event seem remains in the present followed by a past infinitive, as in text. Did seem would transfer the estimate to Bilbo.TN1 I do not think he had time for anything more than relief at feeling the hilt.
Do seem was evidently intended as a remark ‘narrator to audience’, a tiresome trick of which the author was too fond; I have got rid of some of the cases. For that reason (and not for idiom) I propose the emendation:
nor fortunately had the goblins noticed it.
This should fit the available space.
This second query relates to the sixth paragraph of the same chapter, and Bilbo’s discovery that his little sword is an elvish blade: ‘somehow he was comforted. It was rather splendid to be wearing a blade made in Gondolin for the goblin-wars of which so many songs had sung’ (1st ed. page 81; Unwin Books page 64; 3rd ed. page 80; DAA.116; italics mine).
(b) songs had sung. I do not understand the point of this query. The text, besides sounding better, is, I think, perfectly correct English.
If the reader would add to my obligation to him for his attentive reading by stating the nature of his doubt, I should be much interested.
Our sweetest songs are those that tell of saddest thought. If Shelley had written sing for tell, the most cavilling critic could have found fault only in the excessive alliteration. Probably it was to avoid this that the poet used tell instead of the obvious and equally correct sing.
As we see, in this second case, Tolkien rejected the suggested emendation and kept the passage as it was. His reasons for doing so, however, are interesting, as is his rather unexpected citation, apparently from memory, of a line from Percy Bysshe Shelley’s ‘To a Skylark’ [1820] in support for his usage. Tolkien’s supposed lack of familiarity with postmedieval literature has sometimes been exaggerated to the point of parody, and evidence to the contrary is abundant. For example, in his draft for ‘A Secret Vice’, his essay about invented languages (available in The Monsters & the Critics and Other Essays, pages 198–223), he briefly discusses James Joyce and Gertrude Stein (MS. Tolkien 24, folio 44), while in an unpublished essay on translating alliterative verse he at one point compares the Old English poem ‘The Wanderer’ to Wordsworth’s ‘Resolution and Independence’ [1807; better known as ‘The Leech Gatherer’) before comparing the latter in turn to Lewis Carroll’s parody of it, ‘Father William’ [1865], which he much prefers (Ms. Tolkien A30/1, folio 119). In fact, he was of course familiar with the full range of English literature up to about 1830 (thus including the Romantics), the point at which ‘English studies’ terminated when he was a student, and had at least a nodding acquaintance with Victorian and, to a lesser extent, contemporary literature.
Text Note
TN1. Presumably the proofreader had suggested that ‘nor do the goblins seem’ be changed to ‘nor did the goblins seem’, which Tolkien rejects for the reasons stated.
This single page [Ad.Ms.H.S.10], written in Tolkien’s neatest calligraphy, holds a list of all the members of Thorin’s expedition other than Bilbo himself. Its tone would suggest that this was compiled for the Longmans’ ‘school edition’ in 1965/66, but Tolkien paleographer Arden Smith argues that the handwriting has ‘the very distinct look of manuscripts from the 1930s’ (personal communication, April 2011). Although almost all the details given here can be gleaned from the text (and most of them from the first chapter), there are a few details that are new, such as the fact that Dori and Nori are brothers and Ori their cousin, the ages of Fili and Kili (‘about 50’), and the respective girths of Bifur, Bofur, and Bombur.
Dwarves
Thorin Oakenshield, son of Thror son of Thrain [> son of Thrain son of Thror],TN1 King under the Mountain. At the time of the story he was at least 200 [> 195] years old. He had a white beard, a sky-blue hood with a silver tassel, a belt of gold and jewels, and a golden chain round his neck.
Fili and Kili, young dwarves (about 50 years oldTN2), nephews of Thorin. They had yellow beards; their belts were of silver and their hoods were blue.
Balin, oldest of Thorin’s companions, who had escaped with him from the Mountain long ago. His beard was white, and his hood red. He was the look-out man.
Dori and Nori, had purple hoods and silver belts. They were brothers.
Ori, was their cousin. He had a grey hood, and a silver belt.
Oin and Gloin were brothers and the fire-makers. Oin wore a brown hood and Gloin a white hood. Both had golden belts. [added in pencil: Glóin was father of Gimli of the L.R.TN3]
Dwalin was Balin’s brother. He had a blue beard, a dark-green hood and golden belt.
Bifur was a large dwarf with a yellow hood.
Bofur his cousin [> brotherTN4] was even larger. He also had a yellow hood.
Bombur (brother of Bofur) [added in pencil: cousin of <Bifur>TN5] was the broadest and fattest of the dwarves. He had a pale green hood.
Gandalf the Wizard had a long white beard; he wore a tall blue hat, a grey cloak, and a long silver scarf; he had large black boots.
At some later point,TN6 Tolkien underlined the word tall in pencil and added a brief (pencilled) note at the bottom of the page about Gandalf’s hat:
a pity, because in fact I never imagined his hat as being ‘tall’ —only it had a ‘crown’ rather <?point> with a wide brim.
Two sketchy attempts to depict the hat follow, each with a wide brim and low crown – in fact, very like the hat worn by the berg-geist (‘mountain-spirit’) in Joseph Madlener’s postcard (reproduced in The Annnotated Hobbit on the bottom of Plate V), except that Madlener’s kindly old mountain-spirit’s hat has a flat crown while both of the sketches here for Gandalf’s hat have pointed crowns. Contrast this with the two pictures Tolkien drew circa 1936–37 showing Gandalf either approaching or standing at the doorway of Bag-End (Plate III), both of which show a tall, pointy hat with a relatively modest brim; the same hat with a somewhat broader brim can be seen in ‘The Three Trolls are turned to Stone’ (Plate V [top]).
Text Notes
TN1. Here Tolkien originally wrote son of Thror son of Thrain, then encircled ‘Thrain’ in pencil and marked it to be placed where ‘Thror’ stood in the line. It is remarkable that even at this late date Tolkien either could not remember or still remained undecided about which was the father and which the grandfather, Thror or Thrain. This strongly supports the assumption that the earlier back-and-forth between the names was the result of inattention, not indecision; it seems as if his instinct was to make Thror the father, yet he kept changing this back to Thrain as the father in deference to the ‘historicity’ of the established genealogy.
TN2. While this seems in keeping with how the youngest members of Thorin & Company are portrayed in The Hobbit, it contradicts the information in the genealogical chart in Appendix A of The Lord of the Rings, which gives Fili’s age as eighty-two and Kili’s as seventy-seven. See Text Note 3 on page 477 for more on what constitutes a ‘young dwarf’.
TN3. L.R.; i.e., The Lord of the Rings.
TN4. The word ‘brother’ is faintly pencilled above ‘cousin’; see Text Note 34 on pages 322–3 for the shifting relationships between Bifur, Bofur, and Bombur throughout the writing of The Hobbit. Interestingly enough, the original reading of ‘cousin’ here is correct, according to the published text, while making Bofur Bifur’s brother reverts to an earlier conception which is more in keeping with the tradition of rhyming brother’s names (Fili/Kili, Balin/Dwalin, &c.).
TN5. The very faint words ‘cousin of <?Bifur>’ are pencilled above ‘brother of Bofur’; the last word is nearly illegible but can hardly be anything but ‘Bifur’ in the context; in any case, ‘brother of Bofur’ seems not to have been cancelled. Unlike the contemplated change recorded in Text Note 4 above, this addition would not have contradicted the published book, which agrees that Bombur and Bofur are brothers and that Bifur is their cousin. See Text Note 34 on pages 322–3 for more on these relationships.
TN6. Arden Smith suggests that this pencilled addition was much later (‘clearly decades later’ —op cit); I suspect it dates from the Longmans/Unwin period, whatever the date of the list itself.
By far the largest proportion of this ‘Seventh Phase’ material (thirteen pages out of a total of eighteen [Ad.Ms.H.S.1–5 & 11–18]) concerns runes and scripts. In it Tolkien gives multiple drafts of the various rune-systems, followed by two tables of tengwar letters, one arranged in a dwarven mode and the other adapted for use with Modern English.
The first of these charts exists in three variant states [Ad.Ms.H.S.4, Ad.Ms.H.S.3, and Ad.Ms.H.S.1], one of which is labelled
. Futhorc .
A Runic Alphabet.
The ‘futhorc’ is the name generally given to the Old English runes; like the modern word ‘alphabet’ (alpha, ‘a’ + beta, ‘b’), the name is composed of the letters symbolized by the first few runes: f+u+th+o+r+c. Unlike our modern English letters, which derive from Latin, each rune not only stood for a letter but (as Tolkien notes below) had a name which began with that letter; i.e., the name of the 
rune, which stood for ‘th’, was thorn.
[Table I: Dwarven Runes]
[Ad.Ms.H.S.4]
Runes used by Thorin & Co.TN1
There is no Q. For qu cw is used.TN2 The letters represented in English by two letters as ea, oa &c. can also be written with two, as 
for 
. But 
are always used for th, ng, ee, oo. In Runes the same letter is rarely written twice running: either once or the letter is dotted. Thus 
or 
for ‘letter’. 
is used for i and j and 
for u and v. The special shapes 
[added: joy] for j and 
for v TN3 can be used but are not counted in the alphabet. A letter sometimes used is 
or 
= eo, io as in 
= George. It is counted as a combination of I + O and called ‘ice-ox’. The uprights of runes are sometimes joined: as 
= ME. The spelling of words by Thorin & Co is often phonetic: thus 
(dor) for door.
Each rune has a name – a word of one syllable, beginning with the sound represented; except in case of 
where the name has 2 syllables and the sound comes in the middle. Many of these names are trees: thorn, pine, birch, elm, yew, oak, ash.
The text continues onto the verso of the same sheet [Ad.Ms.H.S.5], which presents a second, complementary runic chart.
These runes are very like, but not the same as the Anglo-Saxon Runes. The Anglo-Saxon runes vary a good deal, but were generally like this:
[Table II: The Futhorc]TN4
[Ad.Ms.H.S.5]
The dwarfsTN5 only used Runes for special purposes. For writing they used the Elvish alphabet — a bit of which may be seen on one of the pots of gold in picture of dragon.TN6 It is more complicated but looks better. Here is a bit.
The rest of this page is devoted to a tengwar text of Bilbo’s Contract; compare this neat copy with the freehand style of that of the Frontispiece.
‘Copy of the original of the letter ([translated>] transcribed on p. 39).’TN7
Text Notes
TN1. The last two runes in this line, oak and ash, are circled in pencil by Tolkien and marked to be moved to the front of this row (the extreme left), which would change the sequence of the Fourth row to oak, ash, yew, ear, &c.
TN2. Added above the line in pencil here: ‘or 
quill’.
TN3. At this point Tolkien pencilled a ‘v’ above the letter but did not complete the word; on one draft of this chart [Ad.Ms.H.S.1] this rune had been identified as ‘v. vane’.
TN4. Originally Tolkien wrote two words for eh; eoh and eh, as well as two letters above, eo and e, but in each case he cancelled the first entry (which is already represented by the rune 
in the row above), leaving only eh for the rune’s name and e for its equivalence. As Arden Smith points out (personal communication, April 2011), these are dialectic variants on the same word, with eh being the Mercian and eoh the West Saxon words for ‘horse’ (as in the names Eomer and Eowyn). In the two other versions of this chart [Ad.Ms.H.S. 1 & 3], this rune instead represents the word ‘elm’. I am grateful to Arden Smith for helping me recover the obliterated readings.
TN5. The usage of aberrant form ‘dwarfs’ instead of the usual ‘dwarves’ is Tolkien’s.
TN6. Note that this picture was not included in either the Longmans’ School Edition nor the Unwin Books trade paperback edition, which would have made this reference rather murky to that audience.
TN7. The transcription appeared on page 39 of the first and second editions, and page 38 of the third edition; in the 1966 Unwin Books edition it was on page 27.
Arden Smith informs me (personal communication, April 2011) that Tolkien’s use of the tengwar here in this version of Thorin’s Letter bears strong affinities to his usage in the Ransome letters (see Appendix IV); hence it’s reasonable to assume that it dates from the same time (December 1937).
Continuing his account of dwarven writing systems, and directly following upon the runic material, Tolkien gives the following account of the dwarven use of elven letters. According to Christopher Tolkien, the first half of what follows [Ad.Ms.H.S.11–14 sequence] is written in blue ink; the second half [Ad.Ms.H.S.15–18], which focuses on using these letters to write modern English, is in black ink.
For books or documents or anything written with pens the better kind of Dwarves (as most other folk of the period) used the running form of the ‘Fëanorian’ (or Second Elvish) Alphabet.TN1 As this is a little complicated it was transcribed into our ordinary letters in ‘The Hobbit’ (though there is a glimpse of it painted on one of the jars in the coloured picture of Smaug’s lair).
Here is a copy of the original letter (in Thorin’s rather free and bold style) that is transcribed on p. 39.TN2
In a more careful way (as in making a book) it would have looked more like this.
The ‘book style’ tengwar version of Thorin’s letter fills rest of this page [Ad.Ms.H.S.11].
Fig 4
The following chart and its accompanying text takes up the whole of Ad.Ms.H.S.12.
[Table III: The Tengwar (Dwarven Mode)]
[Ad.Ms.H.S.12]
Adaptation of the Fëanorian Alphabet as used by Dwarves (and others) for the common or ordinary language: ——
| 1. | 2. | 3. | 4. | |
| p-row |  |  |  |  |
| p | t | ch | kTN3 | |
| 5 | 6 | 7 | 8 | |
| b-row |  |  |  |  |
| b | d | j | g | |
| f-row |  |  |  |  |
| f | th(thin) | sh | [x] | |
| v-row |  |  |  |  |
| v | th(this) | zh(French j) | [<yogh>] | |
| m-row |  |  |  |  |
| m | n | n˜(ny) | ng | |
| w-row |  |  |  |  |
| w | r | y | / | |
| odd-row |  |  |  |  |
| h | l | r | s | |
 | ||||
| z |
The letters ‘x’ (f-row 4) and ‘
’ (v-row 4) which had represented guttural sounds no longer heard in the Common Language, were no longer used. The Dwarves occasionally used 
for 
‘h’.
There were originally no vowel-letters[,] the vowels being shown where needed by various marks over or under the consonants: as 
, 
, 
‘tin, tan, bull’. But in writing the common language new letters for vowels were made — but there was a great deal of variation and disagreement about how to use them.
Here is the system used by Thorin (and others).
The letter 
, being no longer used for a consonant, was used for the vowel a.
 =a |  =o ((before letters of column 4 ( etc)  ))TN4 |
 =e |  =u |
 =i |
The Dwarves, however, as a rule used this alphabet to spell what they said phonetically or (more or less) how they pronounced it. They need not be imitated! But they needed a few more letters.
or 
for a in ‘back’; 
= a in ‘father’.
for u in ‘butter’; 
= u in ‘bull’
or a dot underneath for a vague or murmured vowel.
Length of vowel could be marked 
: so 
= father.
The little r 
they used when it was not pronounced (so 
: ‘farther’), but 
when it was: so 
: merry. This 
they used like a vowel for er, ir, ur as in fern, third, burst, butter.
There were various other tricks and abbreviations. S was often written 
or 
as 
= ks, x; 
= ‘bats’; and s, z often 
at the end of a word, as 
= ‘sums’. n or m were usually written 
above a consonant if one followed:TN5 as 
= camp. u or w could similarly be written 
and i or y as 
, but in this case they were put over the letter that preceded. As 
= ai, ay, 
= au, aw; 
= kw = qu.
Some common words were shortened (like our &):
so 
= and; 
= of; 
= of the; 
= the.
To mark Capitals either 
was written underneath, or the downstroke was doubled: so 
= I; 
= D.
For punctuation marks (when any were used) you will find:
as a comma; 
as a full stop; 
or 
as colons (or semicolons).
Exclamations can be marked by 
before or 
after or both.
Questions can be marked by 
before or 
after or both.
For brackets
are used (if anything).
There were no numerals, and if any were required the letters of the alphabet with a dot were used in order except that 
(which had no other use) = 9, and 
= 0.
dwarven numerals (Fëanorian)
But the numerals of the older or First Elvish Alphabet were sometimes used
dwarven numerals (Rúmilian)TN6
Thorin’s letter is very abbreviated and “phonetic” and actually goes like this.
and so on. But Bilbo was used to that sort of spelling.
The rest of this page, almost half of Ad.Ms.H.S.14, is left blank.
As noted above, the second half of this material on the use of Elven letters [Ad.Ms.H.S.15–18] is written in black ink. Unlike the preceding material, instead of looking at dwarven usage it focuses on using Fëanorian script for modern English.
[Table IV:] Fëanorian applied to English.TN7
[Ad.Ms.H.S.15]
Additional Consonantal Signs:
h. 
l. 
r. 
s. 
z.
is only used where pronunciation is consonantal.
is used where [r] has become vocalic murmur or been absorbed. 
ar. is thus used for 
in part. better is spelt 
. 
spelt er, ur, ir in E.[nglish] is represented by 
: as 
= first.
is [cancelled: only] used when followed by w, j [added: or prec. by n] as in 
= swing; 
[added: = dance][;]
= suit; and occasionally when otherwise convenient, esp. finally when an inflexion. In such cases [s] is often reduced to a mere hook 
, and [z] to 
. Thus 
or 
= once; 
= dance; but 
= hats; 
= roses. 
is rarely used for 
. ks = 
, ps = 
[,] ts = 
[.]
A homorganic nasal is usually represented by 
above the following consonant 
= ns, 
= nt, 
= nch, 
= mp, 
= nk, &c. The sound [j, i] is represented usually by 
above the preceding sound: thus 
= assume; 
= tune; 
= oi, 
= ai, 
= ei, &c. The sound [w, u] similarly is represented by 
: as 
= quick; 
= au (ou), 
= ou &c. But initially, between vowels, and at the beginning of distinct elements in compounds the full signs 
[j] and 
[w] are used. They can be used in all positions.
When n, m, l are vocalic as in button, bottom, bottle they are under-dotted: as 
, 
, 
. The same under-dot or ‘vocalic sign’ is used occasionally when the vowel in other cases is reduced to a very slight murmur. Thus in quick ‘colloquial’ writing the reduced atonic pronunciation of such words as at [æt]; of [ov], and [ænd] > 
, 
, n.d is represented thus: 
Note where the nasal over-line is used an under-dot refers to this:
thus: torrent [torn.t] is written 
The underdot is also used to mark 
=[o] as a vowel, when it precedes 
and sometimes before other letters of this series. Thus 
= [wo∫] wash, not [wzˇ].
Special abbreviations are 
= the (for 
or 
). 
= of for 
. 
= of the. In colloquial writing the unstressed indefinite article a, an is often written as part of the next word by underdot : thus 
= a man, 
an ostrich. Unstressed and is written 
= nd. This in quick writing is often reduced to 
.
Other rarer abbrev. 
‘equals’ = 
‘therefore, and so’ 
‘then, next’
Punctuation. Capitals are seldom used. When large initials are required underlining, encircling, or larger writing is used. 
= John. or 
or 
. Brackets 
or 
. The latter are also often used as quotation marks, when no query or exclamation is involved
Full stop 
Comma not much used; when required 
= Tom, the piperTN8
Colon 
Semicolon 
Exclamation 
often also placed before in forms 
Question 
in which case no quotation marks
are used.
The vowels.
The old sign 
originally consonantal is used as [a]
The other vowels are denoted by special signs outside the series, except for vowels developed from [r] which are denoted 
: see above. 
is used = [a] in 
[ai] = diphthong in wine; 
[au] = diphthong in town; before r. as in 
[pat, par.t] = part; when long as in 
= father. The sound [æ] is properly denoted 
(with i-dot), but since this sound is normally defined by position & absence of length the dot is not used much in colloquial writing —esp. not in cases where pronunciation varies between [a¯] and [æ] as in 
[added above the tengwar in neat ink: glass]: here neither dot nor length-sign are usual.
The mark of vocalic length is 
as 
[kam] calm.
Other vowel signs.
[i] 
[e] 
[a] 
[æ] 
[o] 
[u] 
[
] 
or underdot 
or 
[
] 
[y] 
[œ] 
The following is crowded in along the right-hand side of the chart:
<
> which in E.[nglish] is always product of er, ir, ur is written 
.
is treated a pure long [o] & written 
as in 
= all; but or is written 
as in 
= port.
A second vowel chart, denoting superscripts (diacriticals) to indicate vowels, is added below the first, in pencil:
i 
e 
o 
u 
a 
æ 
œ 
y 
In the end, none of these charts and tables made it into the Longmans or Unwin Books edition, although Tolkien did write a note for that edition explaining the runes used in The Hobbit. This prefatory note appears in The Annotated Hobbit (DAA.27) and has become a standard feature of later editions – e.g., the third edition Hobbit (page [8]), the Fiftieth Anniversary Hobbit (pages 9–10), and the Seventieth Anniversary Hobbit (pages 1–2). Perhaps Tolkien did not complete the material in time for it to be included in the book, and another appropriate opportunity did not come – a pity, given how eagerly so much information on Tolkien’s invented scripts would have been welcomed in the mid-sixties by students of his invented languages. Or perhaps it was felt disproportionate to include a lengthy essay devoted to a writing system that did not actually appear in that edition, since the picture of Smaug with Thror’s cup – the only place where tengwar is used within The Hobbit – did not appear in either the Longmans or Unwin editions.
TEXT NOTES
TN1. So called to distinguish it from Tolkien’s older script, the Alphabet of Rúmil, which he had created as early as 1919 and developed during his years at Leeds. For a history of this writing system and several texts in Rúmilic, which looks more like a Dravidian or southeast Asian script than anything in the European tradition,† see The Alphabet of Rúmil, ed. Arden R. Smith, Parma Eldalamberon XIII, pages 3–90.
† for example, Rúmilic tends to be written in columns from top to bottom, rather than in rows like most European scripts.
TN2. This seems to be a reference to the version given as plate two of the frontispiece, since the description ‘in Thorin’s rather free and bold style’ fits that document much better than it does the careful ‘book style’ of Ad.Ms.H.S.11 (page 960) or the inscription style of Ad.Ms.H.S.5 (page 904).
TN3. Tolkien originally wrote ‘g’ here, but cancelled it and replaced it with ‘k’.
TN4. Although Tolkien writes ‘letters of column 4’, the tengwa he gives is in fact the top letter in column 3, and the latter seems to reflect his intent.
TN5. “In that case 
are used for s, z.” [Marginal note by Tolkien, marked for insertion at this point].
TN6. Tolkien has pencilled a third Elvish numeration system in the top margin of Ad.Ms.H.S page 17, but unfortunately it is too faint for me to reproduce with any confidence. Enough of it can be made out, however, to show that it uses the tengwar letters, as in the first numeration system shown here, but it does not assign the same numerical value to the same letters.
TN7. The usual diacritical in ‘Fëanorian’ is missing here in Tolkien’s original.
TN8. This is an allusion to ‘Tom, Tom, the piper’s son; stole a pig and away he run’, an old nursery rhyme (Baring-Gould #126; see The Annotated Mother Goose, pages 104–5).