"It is thus seen that not only the present work but all the history of the Halifirien and Elendil's tomb arose from Mr. Bibire's brief query.
"This is a convenient place to notice a stage in the development of the story of Elendil's tomb that was not mentioned in Unfinished Tales. There is a rejected draft page for the passage recounting the definition of the bounds of Gondor and Rohan by Cirion and Eorl, which scarcely differs from the text printed in Unfinished Tales until the paragraph beginning: ‘By this pact only a small part of the Wood of Anwar....' (UT:306). Here the rejected text reads:
By this agreement originally only a small part of the Wood west of the Mering Stream was included in Rohan; but the Hill of Anwar was declared by Cirion to be now a hallowed place of both peoples, and any of them might now ascend to its summit with the leave of the King of the Éothéod or the Steward of Gondor.
For the following day after the taking of the oaths Cirion and Eorl with twelve men ascended the Hill again; and Cirion let open the tomb. "It is fitting now at last," he said, "that the remains of the father of kings should be brought to safe-keeping in the hallows of Minas Tirith. Doubtless had he come back from the war his tomb would have been far away in the North, but Arnor has withered, and Fornost is desolate, and the heirs of Isildur have gone into the shadows, and no word of them has come to us for many lives of men."
"Here my father stopped, and taking a new page wrote the text as it stands in Unfinished Tales, postponing the opening of the tomb and the removal of the remains of Elendil to Minas Tirith to a later point in the story (UT:310)."
Appendix: The Eldarin numerals
The following text has been removed from the entry for the river-name Levnui (S. ‘fifth') above to this appendix.
The stems of the Common Eldarin numerals (which up to 12 agree closely in the derived languages) were: 1 ‘single' (non-serial) ER; ‘one, first of a series' MIN. 2 TATA, ATTA. 3 NEL, NEL-ED. From 3-9{51} the stems were dissyllabic (Note 1) (triconsonantal, though two of them had no initial consonant, as was not infrequent in Common Eldarin in this pattern): 4 kan-at. 6 en-ek(w) (the (w) only appears in Quenya). 7 ot-os. 8 tol-ot.{52} 9 net-er. 10 kwaya, kway-am. 11 minik(w). 12 yunuk(w).{53} 5 is omitted because it is exceptional. It had the stem lepen, and a supposed variant lemen (but see further below) neither of which ever appeared without the third consonant.
The numerals, as is usual, are mostly not referable with certainty to other stems or bases. The form min is probably the same in origin as MIN that appears in words applying to isolated prominent things, such as steeples, tall turrets, sharp mountain peaks, minya ‘first' thus meant eminent, prominent', cf. Q. eteminya ‘prominent'; also minde ‘turret', augmented in mindon ‘lofty tower', minasse, S. minas: ‘fort, city, with a citadel and central watch-tower'. ‘Five' was no doubt a special number primitively in peoples of elvish/human shape, being the number of the fingers on one hand. Thus lepen is without doubt related to the stem LEP ‘finger' (Note 2). It is also certain that 10 kwaya, and kwayam (-m being also of plural origin), is related to base KWA (kwa-kwa, kwa-t) ‘full, complete, all, every", and meant ‘all, the whole lot, all the ten fingers'.{54} But already in Common Eldarin the multiples of three, especially six and twelve, were considered specially important, for general arithmetical reasons; and eventually beside the decimal numeration a complete duodecimal system was devised for calculations, some of which, such as the special words for 12 (dozen), 18, and 144 (gross), were in general use.{55} But since this appears to have been a relatively late development (only begun after the Common Eldarin [?Period] except for the word for 12),{56} the vague similarity of nel(ed), e-nek-we, net-er are probably not significant.
In Common Eldarin the full forms with ómataima (long or short){57} were employed as cardinals: as Telerin canat, Sindarin canad 4 < kanata. In Quenya the second vowel was syncopated as usual with short unstressed vowels following a stressed vowel of the same quality: hence Q. kanta 4 < kanatā. For 5 Telerin had lepen, S. leben. In Telerin final n (< m, n) was not lost, but it was lost in Sindarin; it is therefore probable that in Common Eldarin *lepen had assumed the form lepene with a final vowel modelled on the other numerals. The Quenya form is lempe. This does not support the view that 5 had in Common Eldarin an alternative stem lemen. In pre-record Quenya the sequences pm, pn, tn, kn were frequently reversed{58}—a process assisted by the frequency with which nasal (homorganic) infixion competed with the suffixion of n, m in word formation, and also by the severe phonetic changes which overtook the voiceless stops before nasals (Note 3). So lepene > lepne would yield lempe without need to substitute m. See further under Ordinals.
The ordinals in Common Eldarin appear to have been formed by addition of adjectival -yā to a stem in which the second vowel was absent. Not by syncope, but according to the primitive modes of derivation from bases. In Quenya the ending -ea was generalized for 3rd, 4th, 6th-9th inclusive. It was the natural form for Quenya in 3rd, 4th, 6th, 9th, and ousted the oya proper to 7th, 8th.{59} The Quenya forms were: 1st minya; 2nd tatya (Note 4) early replaced by attea; 3rd nelya, also neldea; 4th kantea; 5th lemenya (the usual form; lempea only appears in late Quenya); 6th enquea; 7th otsea; 8th toldea;{60} 9th nertea; 10th quainea. The Sindarin forms were cardinal 1 mîn, er; 2 tâd; 3 nêl; 4 canad; 5 leben; 6 eneg; 7 odog{61} (the historical form odo < otoso occurred in Doriathrin according to the grammarians); 8 toloð;{62} 9 neder; 10 pae. The Sindarin ordinals were mainly formed with suffixed -ui, derived from ō-ya, ū-ya (which were present in 7th and 8th), and generalized as a suffix in these and other adjectives. 1st mein, main (from minya, only used in senses ‘prime, chief, pre-eminent', etc.), minui; 2nd taid (only used in senses ‘supporting, second in command', etc.), tadui; 3rd neil, nail (late S. nelui); 4th canthui; 5th levnui; 6th enchui; 7th othui; 8th tollui;{63} 9th nedrui; 10th paenui (Note 5). Other occasional forms are 6th enecthui, with -thui deduced from 4th, 7th, 8th; 7th odothui. othui is the normal and older form, and is directly derived from C.E. otsōya.
51
As first typed, this read "From 3 onwards".
52
Marginal notes against tol-ot show Tolkien experimenting with making the form tol-oth.
53
For the most part, this list of Common Eldarin numerical stems accords with the numeric system stipulated or implied by evidence in The Etymologies and The Lord of the Rings: cf. Etymologies entries ERE- ‘be alone, deprived', MINI- ‘stand alone, stick out’, AT(AT)- ‘again, back’, TATA-, TAT- *'two, double, NEL-, NÉL-ED- ‘three', KÁNAT- ‘four', LEP- (LEPEN, LEPEK) ‘five', ÉNEK- ‘six', OT- (OTOS, OTOK) ‘seven, TOL-OTH/OT 'eight', NÉTER- ‘nine', KAYAN-,
KAYAR- ‘ten', MINIK-W- *'eleven', RÁSAT- ‘twelve'. The two noteworthy exceptions are the stems for 10 (kwaya(m)) and 12 (yunuk(w)), but cf. KWAT-*'full' (V:366), and with yunuk(w) (yielding, presumably, Q. *yunque), cf. YŪ-‘two, both' (V:400).
54
As first typed, this sentence began "Less certain is it that".
55
As first typed, the special words included those for "twelve, eighteen, and the multiples of twelve, 24, 36-144" (i.e., it seems, the multiples of twelve occurring in the range from 36 through 144). A note written in the top margin of this sheet reads: "though for general purposes the numeral names were decimal in origin, special names were devised for multiples of 6".
56
As first typed, this parenthetical remark read: "(only completely carried out in Quenya)".
57
The ómataima is a vowel (óma) of the same quality as the sundóma or base-vowel that is added to a root or stern as an extension (taima). Cf. ómataina (of the same meaning), XI:371, 417; and the bases OM- *'voice' (V:379) and TAY-'extend, make long(er)' (V:391).
58
As first typed, the list of frequently reversed sequences was "pm, pn, pr, pl / tn, tr, tl / kn, kr, kl".
59
That is, the proper forms for 7th and 8th, which had base and stem vowels in o, historically ought to have ended in -oya.
60
The form toldea is an alteration on the typescript from toltea.
61
A prior, deleted version of the entry for 7 reads: "odo (later odog with g from eneg)".
62
Altered on the typescript from toloth.
63
Altered on the typescript from tolthui.