Essekenta Endamarwa - Names from
The Return of the Shadow, The Treason of Isengard and The War of the Ring

Abbreviations

Dor. - Doriathrin
Ilk. - Ilkorin
N. - Noldorin
ON - Old Noldorin
Q. - Quenya

Let - The Letters of J.R.R. Tolkien, quoted by the letter number
LotR - The Lord of the Rings
LR - The Lost Road
LT1, LT2 - The Book of Lost Tales 1, 2
MC - The Monsters and the Critics
MR - Morgoth's Ring
PM - The Peoples of Middle-earth
RC - The Lord of the Rings - A Reader's Companion
Rgeo - The Road Goes Ever On
RS - The Return of the Shadow
SD - Sauron Defeated
Silm. - The Silmarillion
TI - The Treason of Isengard
UT - Unfinished Tales
WJ - The War of the Jewels
WR - The War of the Ring

PE - Parma Eldalamberon
GL - Gnomish Lexicon (PE11)
QL - Qenya Lexicon (PE12)
VT - Vinyar Tengwar


Navigation:
    Return of the Shadow
    Treason of Isengard
    The War of the Ring

If we seek to recapture what they had forgotten,
and examine each of the original elements in turn,
it must be rather for the pleasure of the hunt than
 in hope of a final kill.

(J.R.R. Tolkien ~ Medium Ævum III 2 p. 95)

√TĂR > tāra, tall, ups... sta...
(PE17:186)
 

Introduction

The aim of this article is to compile the Elvish names of persons and locations from The Return of the Shadow, The Treason of Isengard and The War of the Ring, volumes VI, VII and VIII of The History of Middle-earth, and to analyze them linguistically. These volumes deal with earlier drafts of The Lord of the Rings and thus nearly all of the names presented here are external predecessors of those later ones which have found a way into the book.
For the most part I have ignored those names that have not changed and are identical to those in The Lord of the Rings (like Aragorn, palantír.. etc.), unless analyzing them was helpful to understand the earlier forms.
All words are presented roughly in the order of their occurrence in the three books, but sometimes it was suitable to summarize several forms sharing the same elements or referring to the same place or person. As I also intend to keep the possibility of using this article as a reference for particular forms, repetitions were inevitable, since a couple of elements occurs over and over again.
Tolkien's reconstructions are not asterisked, just my own. Whenever roots were used in the analysis, they are from The Etymologies (LR:347-400, VT45/46), unless differently glossed. The work on this source was begun by Tolkien simultaneously with the writing of The Lord of the Rings and is thus highly important in the context of this article.
Some interpretations are easy and unambiguous, which is why they are given without much commentary. Others remain a riddle even with the help of The Etymologies. Clearly, Tolkien had no necessity to note the root and origin corresponding to every single name in his legendarium, while there might still be unpublished linguistic material (from this stage or another) which could throw a light at some of the names.

The main problem is that Tolkien was perfectly free to invent new roots and origins for the new names as he went along. An analysis in terms of the known roots or elements might prove totally wrong, but no other way is left to us. So, as Tolkien puts it in the introduction to an analysis of his own, it must be rather for the pleasure of the hunt than in hope of a final kill.

Another problem in analyzing all these names is the lack of knowledge about the actual language they belong to. At this stage of mythological conception the Elves of Beleriand spoke Ilkorin, while the external predecessor of Sindarin, Noldorin, was spoken by the Noldor that came to Middle-earth (see e.g. LR:177). In its development from Old Noldorin it was also influenced by Ilkorin and both languages are overall very close phonetically, which makes them difficult to distinguish. Such a distinguishing is not really important in order to understand the names, but it may lead to a wrong picture of either language.

Earlier discussions of some of the forms were led on the Lambengolmor mailing list and in the Quettaron minaþúrie thread on the site Aglardh.


 

Names from The Return of the Shadow

Tolkien started to write a sequel for The Hobbit end of 1937 and began the work several times all over again, until he came to the Mines of Moria by the end of 1939 (RS:11,461). Most of the linguistic material is found in the third phase, which began some time in 1939 (RS:309).

● Elberil (RS:68)
This is a variant of later Elbereth and it contains of course el- 'star' N only in names (EL-). The latter element of Elbereth is bereth from Barathī *'Queen' with i-affection (BARATH-). Likewise, #beril *'lady, queen' is probably derived from the shorter stem BAR- yielding words for 'save, rescue', but: original significance probably 'raise'. The stems BAR-, BARATH- and BARÁD- (the last produces words like 'lofty', 'sublime') are all cross-referenced in The Etymologies, being apparently related. Compare also Adunaic bār 'lord' or 'king' (SD:428,437).
For the feminine suffix -il compare brannon 'lord', brennil 'lady' (BARÁD-).

● Ilverin (RS:180,187)
● Neldorín, Elberin, Diarin (RS:187)
The first name appears instead of Dairon in a poem about Beren and Lúthien entitled Light as Leaf on Lindentree, the others are added in pencil on the margin.
Ilverin already occurs in the Book of the Lost Tales, where he is an elf from the 'Cottage of Lost Play'. In English he is called 'Littleheart'. The first form written was Elwenildo, apparently containing Elwen (n) 'heart' (QL:35,52) from LEFE, whence also Q. lepsa 'liver'. Some Gnomish forms are also given: Ilfing, Ilfin, Ilfiniol 'little one, little heart' with ilf 'heart' (GL:50) and #-ing, #-in, #-iniol being diminutive suffixes (the last two appear to be adjectival, however).
Since nothing similar to these early elements appears in later sources (Q. lepse means 'finger' (LEP-, LEPET-)), we might assume that Tolkien liked the sound of this name and took it over, intending to reinterpret its meaning.
Neldorín may be clearly referred to Dor. neldor 'beech' (NEL-, NÉL-ED-). The suffix -ín may have agentive sense (cf. Duil Rewinion, Melin below), but curiously there is a long vowel (which might also be a misreading). Alternatively -in may be an adjectival ending, compare for instance Dor. ngorthin 'horrible' (ÑGOROTH). Finally, it could also be the Doriathrin cognate of N. ind, inn 'inner thought meaning, heart' (ID-); thus either *'one of the beech tree' or *'beech-heart'.
Elberin could be the masculine counterpart to Elberil above, i.e. containing Dor. el 'star' (EL-) and #berin *'lord' (BAR- + agentive #-in). But #berin could also be the adjective 'bold' cognate to N. beren (BER-) or even lenited perin *'half, divided in middle' (PER-), if such a sound change occurred in Doriathrin. In this last case, one would have to interpret the name as *'half-elf' (cf. the remarks on the confusion between 'elf' and 'star' in WJ:378), but there is no hint that Dairon or Daeron was conceived to be a half-elf in the legendarium. Compare also the Doriathrin name El-boron (BOR-), later removed.
Diarin looks similar to Dor. Dior 'successor' (name of Thingol) which is derived from ndeu̯ro < NDEW- 'follow, come behind'. Assuming that Diarin is derived from the same stem, the change -eu̯- > -ia- rather than -eu̯- > -io- is difficult to explain, however (unless it is a misreading). The ending #-rin could be now a variant of -in, just as we see e.g. nathron 'weaver, webster' (NAT-) beside #faron 'hunter' (SPAR-). But also the meaning *'following heart' would suit Dairon's role in the legend, so perhaps #diar might be an adjective from NDEW- or itself mean *'follower'?

● Rimbedir >> Padathir (RS:194,198,361)
● Du-finnion (RS:361)
● Ethelion 'Peregrin' (RS:392,395)
These are Elvish names of Trotter, whose character later became Aragorn, but was originally conceived as a hobbit. For a detailed analysis see Lambengolmor #915 and #918.
In short, there are two possibilities for Rimbedir - #rim- 'frequent' (RIM-) + lenited #pedir, an agentive formation from pata- 'walk', as in S. aphad- 'follow' < ap-pata 'walk behind, on a track or path' (WJ:387; compare also PATA in QL:72), thus *'he who walks a lot'; or rhimp 'rushing, flying' (RIP-) + #pedir, thus *'rush-walker' - the latter explanation is closer to 'Trotter'.
Padathir is clearly formed from pata as well, but there are several possibilities for the ending. We know N. †dîr surviving chiefly in proper names (DER-) - here it could be lenited to -ðir and then dissimilated to -þir. Another possibility is the attachment of the ending -ir to the noun #padath 'walking', compare e.g. tirith 'watch, guard' (TIR-), although many examples seem to indicate that -th is preferred after r or l in the stem (e.g. mereth 'feast, festival' (MBER-), meleth 'love' (MEL-), dalath 'flat surface, plane, plain' (DAL-), palath 'surface' (PAL-) and so on).
Another possibility may be a connection with TIR- 'watch, guard', perhaps via *patat-tir or *padad-tir, but then meaning *'Scout, Ranger' rather then 'Trotter'.
Du-finnion most probably means *'dark-haired' < N. dû 'nightfall, late evening' or 'night, dead of night' (DOƷ-, DÔ-) + find, finn 'a single hair' (SPIN-, PE17:17) + name-formative ending -(i)on. Compare durion 'the Dark Elf'. Alternatively we might be dealing with an adjective finnion *'haired, having hair', see note.
Finally, Ethelion must be connected to Q. ettele 'outer lands', ?ettelen or ?ettelea 'foreign' (VT45:13, reading uncertain). Medial -tt- would produce -th- in Noldorin/Sindarin, hence Ethelion *'foreigner', which is the meaning of 'Peregrin'.

● Orendil >> Elendil (RS:197-198,260,270)
● Ithildor >> Isildor >> Isildur (RS:261,271,320)
Tolkien hesitated about the names of Sauron's defeaters and he also considered Valandil (cf. VT46:4, LR:60) for later Elendil.
The first name, Orendil, could mean *'lover of the sunrise/morning' with #ore as Q. óre 'rising', anaróre 'sunrise' (ORO-), Early Qenya ōre 'the dawn, Sun-rise, East' (QL:70) and the suffix -(n)dil which implies 'devotion', 'disinterested love'  (Silm.index) and is usually translated as 'lover', e.g. Aiwendil 'Lover of Birds' (UT:401, NIL-, NDIL-). Compare also the name Orendel from a Middle High German poem probably having connotations with sunrise and morning, being a cognate of Old English Ēarendel, which was Tolkien's inspiration for his character Earendel, later Eärendil (Let:297).
Ithildor contains N. Ithil 'poetic name of the Moon' (THIL-) and the suffix taur < taurā 'mighty' often found in names, as Tor-, -dor (TĀ-,TAƷ-). However, in Isildor Q. Isil 'moon' is substituted, but Quenya usually does not change ā > au >o as Noldorin/Sindarin. So perhaps Isildor should be interpreted as containing the suffix -do as in pl. Hildor 'the Followers' < KHILI (WR:387) or Q. meldo 'friend' (VT45:34) < MEL-; and another name-formative or agentive suffix -r.

● Rhimdath 'Rushdown' (RS:205) < N. rhib-, rhimp, rhimmo 'to flow like a [torrent?]' (RIP-) + dath 'hole, pit' (DAT-,DANT-) or rather 'steep fall, abyss' (VT45:8)
This name is given directly under RIP- with the variant Rhibdath. See also Rhimdad, Rhimdath, Rhibdath 'Rushdown' below.

● Narothal 'Firefoot' (RS:345,347,351) [Gandalf's horse] - N. tâl 'foot' (TAL-)
● Narosîr 'Redway' (RS:433) - N. sîr 'river' (SIR-)
● Narodûm 'Red Vale' (RS:433) - for dûm 'valley' see the discussion of the other names of the 'Red Vale' below
I have put these three words together because of the common initial element meaning 'red' or 'fire', doubtlessly related to the stem NAR1- 'flame, fire' with the derivative narwā 'fiery red' > N. narw, naru 'red'. Compare also Nargalad, Nardol below.
Variation between o and u is a common phenomenon in Noldorin/Sindarin. For instance, u becomes o if the following syllable contains a, e, o or ō [4], but not if this u comes from the consonant w, as it may be the case here (narwā), unless one might take naro- as analogical. But we may also be dealing with o from the vowel u, take a look at the following commentary:
-u- suffix frequent in Q[enya] after el, al, see Q[enya] Structure; cf. kelu, telu, smalu, etc. (VT46:8)
This also applies to Noldorin and Ilkorin, which further often shift u > o; see e.g. N./Ilk. celon 'river' < kelu + n (KEL-) or Celos, derived from kelu- 'flow out swiftly' + -sse, -ssa as in Q. kelussë 'freshet, water falling out swiftly from a rocky spring' (UT:426). And -u- is perhaps also be favoured after r, compare e.g. Q. erume 'desert' (ERE-), Eru 'the One' (UT:305).
Another interpretation might be that the first element is the genitive of *nar 'flame'. Genitival inflections were lost in Noldorin/Sindarin with the final vowels, but may have been preserved medially. This may explain the lack of lenition of sîr and dûm, as well as the usage of the circumflex - the compounds being put together out of earlier expressions like *naro sīre > *naro-sîr, but it does not explain the spirant mutation in Narothal.
A third explanation could be that the first part is actually ?naron, an adjective similar to caron 'red' below. The consonant -n then causes nasal mutation of tâl to -thal (although other examples show lenition instead). In the case of Narosîr it is then assimilated to -ss- > -s-, compare S. besain < 'bread-giver' < *mbassaniē; Q. massánie (PM:404). But this is very shaky indeed and cannot explain Narodûm.
And finally a fourth possibility is an adjective of the form ?narod < narātā, which may explain both Narothal < *narautthal < *narāt-tal and Narodûm < *narod-dûm < *narāt-dûm, but not Narosîr, where one would rather expect *Narothîr, as in N. ethir (ET-, SIR-). But after all, one should not expect that Tolkien had to use exactly the same element in all of these names.

● Beleghir 'Great River' (RS:410,434) < beleg 'great' (BEL-) + sîr 'river' (SIR-), lenited -hir
This is an earlier name of Anduin. It is unclear whether -gh- represents here two sounds (g and h) or just a single back spirant, which for example also appears in Morghul at this stage of the development of Noldorin. Later names for this river are Sirvinya, Andon (see below).

● Annerchin 'Goblin Gate' (RS:416,432)
The first element is most probably #ann 'gate', derived from AD- 'entrance, gate' via adnō > *and, ann. Compare also Early Noldorin ann 'door' (PE13:137).
The latter element looks like orch 'goblin' (ÓROK-) with the plural ending -in. Ilkorin affects o > u in the plural, e.g. thorn 'eagle' > pl. thurnin (THOR,THORON, VT46:19), but also o > e in talum 'ground, floor' > pl. telmin (TAL-, TALAM). Plural -in can be observed in Noldorin as well, for instance naugol 'dwarf' > pl. nauglin (NAUK-). The plural form erchin should be probably seen as uninflected genitive in this context. Alternatively it might be the adjectival ending -in, then it would be literally *'the goblinish gate'.
In The Return of the Shadow Annerchion is found (TI:114) with the ending -ion, on which see the note at the end.

● {Tar} Arad Dain - probably 'High Pass' (RS:416,432)
Christopher Tolkien writes:
I do not understand the reference of ‘Arad Dain (Annerchin)’. My father first wrote Tar and struck it out before writing Arad.
But with Tar Tolkien seems to have begun a word meaning 'high' - see TĀ-, TAƷ- 'high, lofty, noble' and its derivatives; the names of the kings of Númenor mostly start with Tar- 'high'; compare also Tarkilmar 'Westermanton' and Tar as an early name of Celeborn below. This is however, Quenya - Noldorin has forms with a long a, later diphthongized to au > o. But a root TAR- is cross-linked under ƷARA- (VT45:17) and TUR- (VT46:20) in The Etymologies, although it does not appear as an entry. So it seems likely that Arad Dain refers to the 'High Pass' on page 416.
If so, arad may have two explanations - it may be the adjective 'high' (compare later ARA- 'noble', arātā > S. arod (PM:363)). Or it could be a derivative of RAT- 'walk' (whence N. râd 'path, track') with prefixed sundóma meaning 'pass'.
According to this, dain should then mean 'pass' in the former case and 'high' in the latter and is perhaps lenited. The stems TAN- 'make, fashion' and (N)DAN- 'back' do not suit the meaning, but DEN- 'hole; gap, passage' with the derivative N. dîn 'opening, gap, pass in mountains' seems to do. Thus one could explain dain by the sound change *denjā > *dein > dain *'pass'. The suffix -jā is common in forming adjectives, but compare also Q. kirya, N. ceir 'ship' (KIR-) or Q. ranya, S. rein, rain 'erratic wandering' (VT42:13) from primitive *kirjā, *ranjā.
However, dain is also directly attested in Early Noldorin as 'high, noble, divine' (PE13:141), derived from dagná. By the time of The Etymologies we have DAƷ- 'great' (VT45:8) with daʒrā > N. daur 'great, large'. According to the phonology of Noldorin as presented in The Etymologies the back spirant ʒ becomes e as part of a diphthong in such a position, cf. maʒiti 'handy, skilled' > N. moed, but Tolkien wrote later:
Originally the difference between correct Sindarin ae and ai was neglected, ai more usual in English being used for both in the general narrative. (VT42:11)
Arad Dain was put down very early, we do not know how much of The Etymologies was written at that time. In Early Noldorin at least we always see ai as in the very example of dain. Hence, the most probable explanation appears to me: arad *'pass' and dain 'high' with the adjective in the usual trailing position and without lenition.

Now follow some place names containing the colour 'red', best described by Gandalf:
‘Well, anyone who did look at the map,’ said Gandalf, ‘would see that away there stands Taragaer or Ruddyhorn, - that mountain with the red side. The Misty Mountains divide there and between their arms lies the land of Caron-dûn the Red Valley. Our way lies there: over the Red Pass of Cris-caron, under Taragaer's side, and into Caron-dûn and down the River Red-way - to the Great River, and . . .’ He stopped. (RS:419)

All these names also had a lot of experimental sideforms. Let us at first take a look at the river 'Redway' (a tributary of Beleghir >> Anduin), later becoming Ruinnel (see below) and eventually 'Silverlode' or Celebrant.
Caradras dilthen refers to a narrow watercourse in a deep channel on the western side of the Misty Mountains, before the Moria Gate. It was dry, and there was now no water among [the] reddish stones in the bed (RS:445). It appears later as Sirannon, the 'Gate-stream' (ibid., LotRII ch.4).

● Crandir 'Redway' (RS:432)
● Rathgarn >> Rathcarn (RS:433)
● Caradras 'Redway' (RS:433,438)
● {Nenning} (RS:433)
● Narosîr 'Redway' (RS:433) - see above
● Caradras dilthen 'the Little Redway' (RS:462)
● Celebrin (RS:434)
In Crandir, the first element is obviously the result of a syncopation in the stem KARÁN-, cf. N. crann 'ruddy (of face)' (LR:362). The second element may be lenited *tîr, from the stem TER-, TERES- 'pierce', whence e.g. N trī 'through' (VT46:18), prefix tre-, tri. This stem may have produced the word for a way or path (see also Terch Ungol below). It is notable that an Early Noldorin poem given in MC:217 begins with dir avosaith, which, judging by the translation, should mean 'through gloomy places', so that one would have dir *'through' < *DIR, *DER (?).
The next form Rathcarn contains rath 'course, river-bed' (RAT-) and #carn, now a syncopation of the second vowel in KARÁN-, similar to Q. karne; lenited in Rathgarn.
Caradras contains N. caran 'red, blush, red [?part] of face' (VT45:19) or 'red, ruddy' (PE17:36) without syncopation. This name was later assigned to the mountain 'Redhorn', see RAS- in The Etymologies and VT45:10, but before r(h)as must have meant 'way' (in this context see also Ered Nimrath below). The combination -dr- probably stands for -ðr-, compare Maidhros (anglicized Maidros) (MAD-). At least in The Lord of the Rings the mountain's name is Caradhras (PE17:36).
Nenning is difficult to interpret - it also appears later as a river in The Silmarillion, where it is said to contain nen 'water', but the latter element is obscure - maybe related to ING- 'first, foremost'?
The adjective dilthen 'little' looks similar to N tithen 'little, tiny' (TIT-) and is thus perhaps lenited from *tilthen, being derived from a parallel or changed stem *TILIT-.
Finally, Celebrin (substituted in red ink) is plainly the adjective 'of silver' (KYELEP-, TELEP-). However, this was changed to celebren in The Etymologies with a-affection i > e in the final syllable. We know that Ilkorin had no a-affection (compare e.g. N. lhend with Ilk. lind (LIN2-)), so that Celebrin may be Ilkorin.

Now to the 'Red Pass', a passage, where the Misty Mountains may be crossed (in The Lord of the Rings just 'the pass of Caradhras'):

● Cris-caron 'Red Pass' >> Criscarn (RS:419,433)
● Cris-caron, Cris-carron, Cris Caron (RS:433)
All names apparently contain N. criss 'cleft, cut' (KIRIS-)
The adjective 'red' takes again various forms; #carn appears as in Rathcarn. The form caron seems unusual, however. A parallel primitive form *carāna with a long ā could have produced caron by regular Noldorin change (compare NÁRAK- > narāka 'rushing, rapid, violent' > N. narog). On the other hand final -on occurs several times in Ilkorin and three out of four examples are adjectives:

• Ilk. gelion 'bright' (LR:360)
• Ilk. tovon 'lowlying, deep, low' < tubnā (TUB-)
• Ilk. Brithon 'pebbly' (BIRÍT-)
• Ilk. thavon 'carpenter, wright, builder' < stabnō (STAB-)

In his article on Ilkorin [1] Helge Fauskanger suggests a development via a final syllabic nasal -n:

• *stabnō > *thavno > *thavn. > Ilk. thavon
• tubnā > *tobna > *tobn., *tovn. > Ilk. tovon

This seems plausible and one may imagine *carn > *caron as a form parallel to #carn (perhaps more modern). In any case both -n and -on were already quite usual adjectival endings in Goldogrin often used both after final -r, as for instance barn, baron 'tilled, inhabited' (GL:21); farn, faron 'separate, different, strange' (GL:34); marn, maron 'ripe' (GL:56). See also the note at the end.
But a peculiar form is carron with double -rr- medially. It might be a result of assimilation of the combination -rn- as being syncopated *caran- > *carn- > *carr- with the same adjectival ending -on (< -ānā?). However, there are no similar examples, -rn- being a permitted cluster in Noldorin (and Ilkorin).
Assuming that the stem KARÁN- may be an extension of shorter *KAR- (although this clashes with KAR- 'make, do'), carron could be the result of two adjectival endings -rā and -nā with -*-rānā > *-rauna > -ron. Compare WJ:392, where Tolkien discusses KWENE (whence Quendi 'speakers') and KWETE 'speak, utter words, say' as elaborations of KWE; and KWATA (whence Q. quat- 'fill'), KWAN as elaborations of KWA denoting 'completion'.

Then there is the mountain 'Redhorn':

● Bliscarn >> Carnbeleg >> Taragaer 'red horn mountain' (RS:433)
● Rhascaron (RS:433)
The first name obviously contains #carn 'red' once more, but the element #blis- can be hardly related to any attested stem (*BILIS-?) or word. The following form contains beleg 'great' (BEL-), so perhaps #blis- is another derivation from the same stem? Assuming something like primitive b'lessē one could theoretically arrive at *blês >*blîs, but this is rather shaky.
For the next form we can find a direct reference in The Etymologies - Tarag(g)aer 'Ruddihorn' < N. tarag 'horn, also used of steep mountain path' (TARÁK-) + N. gaer 'red, copper-coloured, ruddy' (GAY-).
Rhascaron transparently consists out of ?rhass 'horn especially on living animal, but also applied to mountains' (RAS-,VT46:10), of uncertain reading in The Etymologies (S. rass 'horn' in PE17:36) and caron 'red', for which see above. This word is written on the margins of the page, where the other forms occur; along with 'Caradras = Ruddihorn'.

And finally  follow quite a few forms for the 'red valley':

● Caron-dûn 'Red Valley' (RS:419)
● Carondoom 'the Red Valley', {Carondûn, Doon-Caron} (RS:433)
● Carndoom, Caron-doom, Caron-dûn, Dun Caron 'Red Valley' (RS:433)
● Carondoom >> Dunruin (RS:454,464)
See also Narodûm above. From all these forms one can isolate the word 'valley' appearing in the variations dun, dûn, doon, doom.
The Etymologies give TUB-, tumbu 'deep valley under or among hills', N. tum. This already looks similar and one may guess that the forms starting with d- are lenited, but in fact they are not; or at least not all of them, since initially Dunruin and Dun Caron occur.
Looking further, the stem DUB- in The Etymologies looks related to TUB-, meaning 'lie, lie heavy, loom, hang over oppressively (of clouds)'. TUB- yields e.g. Q. tumna 'lowlying, deep, low', while DUB- yields Q. lumna 'lying heavy, burdensome, oppressive, ominous'. Additionally, there is the stem pair NDUB-, DUB- (<< NDUM-, DUM-) 'lay base, foundation, root, found', N. dum 'root, foundation' (VT45:38), while Ilkorin dûm 'twilight' (DOMO-) differs in sense.
So maybe Tolkien decided to derive 'valley' from DUB- or NDUB-; or perhaps rather 'dark, ominous valley'. This would explain the forms with final -m, but what about final -n? DUN- 'dark (of colour)', N. donn 'swart, swarthy' does not really fit. Was another root change in progress here? Perhaps it is comparable with TALAM- 'floor, base, ground', N. talaf 'ground, floor', but later S. talan 'The wooden platforms in the trees of Lothlórien on which the Galadrim dwelt' (UT:465). According to PE17:52 talan, tâl 'flat space, platform' is still from TALAM- (CE talam-). In VT47:24 Tolkien notes that Common Eldarin final -m, became -n (with the same subsequent development as n in the descendant languages). If this applies here, we would have *(n)dummē or *(n)dumbē > dûm beside *(n)dum- > dûn. Compare also √TAM 'construct' appearing as √TAN in Sindarin by influence of √PAN 'arrange, set in order' (PE17:107-108).
The forms doon and doom are clearly adapted to English orthography, similar to the river Lune; where the actual Elvish name was Lhûn (LotR App.F, RC:773). For carn, caron 'red' see above.
Dunruin seems to contain lenited N. gruin 'ruddy' (ROY2-) at the first glance, but in The Treason of Isengard we can later on find Ruinnel 'Redway' (TI:114), leaving the basic form inevitably as #ruin - see the discussion of Ruinnel below.
Interestingly, we encounter Carn Dûm later as 'the chief fortress of Angmar' (LotRI ch.8, App.A, UT:425), but without any hint on how the name is meant to be interpreted.


Now to the other names:

● Tum Dincelon 'Dimrill-dale' (= 'Vale of Dim Streams') (RS:432,466)
● Nanduhiriath (RS:433)
The Dimrill-dale had been originally put to the northeast of Imladris, before Tolkien moved it to the other side of the Misty Mountains, see also Nen fimred, Nenvithim 'Hoardales' below.
It is easier to interpret the later form first - Nanduhiriath consists apparently out of nan, nand- 'valley' (Let:230, Silm.index) [The Etymologies give NAD- with N. nann, nand 'wide grassland' and Dor. nand 'field, valley'] + N. dû 'nightfall, late evening' or 'night, dread of night' (DOƷ-, DÔ-) + lenited siriath 'rivers' in the collective plural (see SIR-).
The final from Nanduhirion in The Lord of the Rings contains sirion '[great] river' (ibid.). Tolkien explains it as S. nan(d) 'vale' + dû 'dimness' + sîr 'stream' + (i)on which appears in various later-formed names of regions (PE17:37). The ending -ion < √YŎNO 'wide, extensive' is on another occasion said to be a blend with augmentative or male -on (PE17:43), so Sirion 'the Vale or lands about the River Síre' or 'the great stream' (ibid.). Compare note.
The earlier name seems to mean the same with N. tum 'deep valley under or among hills' (TUB-). The former element of Dincelon is perhaps from the stem DEM-, earlier DIM- (VT45:9) with derivatives like Ilk. dimb 'sad', dim 'gloom, sadness', dem 'sad, gloomy'. Compare also Dimbar from The Silmarillion, the first element of which means 'sad, gloomy' (Silm.index). The final -m could be assimilated to the following c- resulting in -nc-, compare Goldogrin dungort, dumgort 'an (evil) idol', dum 'secret, not to be spoken, especially of bad things' (GL:31).
The last element is clearly N./Ilk. celon 'river' form KEL- 'go, run (especially of water)'; hence uninflected genitive: *'valley of a gloomy river', matching the translation.

● Palathrin (Palath = Iris) (RS:432) - the 'Gladden' river
Tolkien wrote in Let:297 about the flower gladden:
[...] the local names Gladden River, and the Gladden Fields, which contains A.S. glædene 'iris', in my book supposed to refer to the 'yellow flag' growing in streams and marshes: sc. iris pseudacorus, and not iris foetidissima
The later name of this flower, ninglor, is translated as 'golden water-flower' (UT:450); consisting out of S. nîn 'wet' < nēnā (Silm:nen, PE17:52) and glor 'gold' (GLAW(-R)-). This earlier name, however, contains no identifiable element meaning 'gold' or 'yellow'. The word palath (PAL-) means 'surface' in Noldorin instead; the stem is glossed 'wide (open)' and produces words like Q. palu-, palya- 'open wide, spread, expand, extend', N. pelio 'spread'. So perhaps palath is the *'wide open flower' with reference to its long and broad petals? Note also that palath also appears with the original meaning in Palath Nenui 'Wetwang' ('wang' = 'field, flat area'), see below.
Here, palath is given the adjectival ending -rin comparable for instance to celebrin 'of silver' < celeb 'silver' (KYELEP-,TELEP-). The absence of a-affection may mean that the word is from Ilkorin (see river Celebrin above).

● Penedrethdulur 'Dimrill-stair' (RS:433)
The first element can be clearly identified with N. pendrad, pendrath 'passage up or down slope, stairway' (PEN-,PÉNED-), here in the unsyncoped form. But pendrad, pendrath contain râd 'path, track' (RAT-) or rath, translated as 'course, river-bed' in The Etymologies (RAT-), later 'street' (UT:461). An explanation of the shift of -a- to -e- in pendreth is difficult. Could it be assimilation to the e-e sequence of pened- or pluralization? Perhaps it is just a slip or misreading.
The second element may be connected with the stem DUL- 'hide, conceal', whence N doll (dolt) 'obscure'; and is perhaps formed with ómataima-extended *dulu- and the adjectival ending -rā > -r, as in taurā 'mighty' > N. taur (TUR-), the sequence u-u then resisting a-affection [4].

● Thanador >> Ulthanador >> Borthendor >> Orothan[ador] (RS:434)
These are earlier names of later Rohan.
An obviously frequent element contained in three of them is #thana. There is a later root THĂN- 'kindle, set light to' (MR:388, PE17:187), but the meaning does not fit at all. Intitial th- in Noldorin/Sindarin can also be produced by ST-. The Etymologies give STAN- 'fix, indicate, decide' with N. thand, thann 'firm, true, abiding' (VT46:16), Q. sanya 'regular, law-abiding, normal'. The meaning 'firm people' referring to the character or attitude would make sense. However, to get a word like thana with a final vowel in Noldorin one needs a stem expansion in -G (*thanaga), compare the very similar N. thala 'stalwart, steady, firm' < ON stalga (STÁLAG-).
Perhaps Tolkien has decided to alter these roots or maybe the final vowel -a survives in the compound medially, as it does e.g. in Magladhûr 'Black-sword' (MAK-), while it is lost in the isolated word finally: makla > N. magl, magol 'sword'.
Hence, Thanador would mean *'land of firm ones' < #than(a) *'stalwart, firm' + dôr 'land, dwelling-place, region where certain people live' (NDOR-, VT45:38).
The second form further contains the prefix ul-. In The Etymologies under ÚLGU- (ÚLUG-) some Easterling names devised by Elves are listed with this prefix: Ulfang, Uldor, Ulfast, Ulwarth and it must mean 'hideous, horrible'. Indeed, we get to know that the Horse-kings have long been in the service of Sauron, although this conception was apparently quickly changed.
Already the next form contains #borthen, probably a derivate from the stem BOR- 'endure' with N. boron 'steadful, trusty man, faithful vassal' and names of 'Faithful Men' - Bór, Borthandos (Borth + handos), Borlas, Boromir. So #borthen could be the adjective *'trusty, faithful'; Borthendor = *'land of faithful ones'. For the formation compare N. malthen 'of gold' (SMAL-).
The last form then seems to contain a prefix oro- *'high', appearing in The Etymologies as or- in orchall, orchel 'superior' (ORO-). As the end of the word is put into brackets, however, we must assume that Orothan may already stand as a name; on the same page is actually written: Rohan = Rochan(dor) 'Horseland'. Then, Orothan would contain the same suffix -an for 'land' (Tolkien translates Rohan < Rochann < S. roch 'horse' as 'Hippia' in Let:144), leaving the basic form #oroth-, perhaps a noun comparable to N. meleth 'love' from MEL- or tirith 'watch, guard' from TIR-. Hence - Orothan *'land of exalted ones'? Or does it simply refer to the terrain - *'high-land', *'rising land'? But see also Early Noldorin oroth 'impetus, haste, speed, rash courage' (PE13:151), which also seems suitable - especially considering the horses and cavalry.

● Rohiroth, Rochiroth 'Anaxippians' (= 'Horse-lords') (RS:440) < Rochir 'horse-lord' < roch 'horse' (ROK-) + kher- 'possess' (Let:144, cf. KHER-) + -hoth 'host, crowd, frequent in people-names as Glamhoth (KHOTH-)
For the loss of h compare Lossoth 'the Snowmen' (LotR App.A), Loss(h)oth (PE17:39), apparently formed with S. loss 'snow' (VT42:18). See also Rochirwaith and related forms below.

● Mellyn, Meldir >> Melin 'friends' (RS:452,462)
In this early draft the Moria Gate opened to the plural of the word 'friend'. The first form is transparently the plural of N. mellon 'friend' (MEL-) by i-affection. The form meldir, also occurring as 'friend' in The Etymologies seems to be identical to its plural.
The last form Melin could be the plural of melen *'beloved, dear' < *melinā or *melēnā, compare the formation of N. thoren (< tháurēnā) 'pp. of thoro- 'fence'' (THUR-). The form melen was written untranslated under MEL- and struck through when the first letter of its plural form was reached (VT45:34), Perhaps it was replaced by melui 'friendly, lvoing, kind'. On the other hand there is already the Quenya adjective melin 'dear', which may occur unchanged in Noldorin as well, also being identical to its plural. A third possibility may be that melin is derived from the same primitive form as Q. melindo 'lover' (m.), thus *mel-indō > N. *melin(n). A similar formation #rewin *'hunter' may occur in Duil Rewinion (see below).

● Erceleb >> Ithil (RS:458,465)
This are earlier names for the substance Mithril.
The first form contains the element er-, on the first glance related to ERE- 'be alone, deprived' (whence N. ereb 'isolated', eriol 'alone, single') and N. celeb 'silver' (KYELEP-, TELEP-). How does 'isolated' or 'deprived silver' make sense? Well, perhaps it is intended to mean *'pure silver'. Compare also the Quenya superlative ĕrĕmelda 'sole dear, dearest of all' (PE17:57).
In Q(u)enya, at least, the element er was also used in counting, e.g. er 'one (and 'the first')' (PE14:49), also Gn. er 'one' (GL:32) so perhaps Erceleb is *'the first silver' or even *'the one = the true silver'?.
In the Qenya Lexicon, however, the stem ERE(N) means 'iron or steel' and the early Qenya names of Angband are Angamandu or Eremandu 'Hells of Iron' (LT1:249). If Tolkien decided to keep this, Erceleb would be the *'silver-iron' or *'silver-steel'.
The second form Ithil is plainly 'moon' (THIL-), but is perhaps rather to be understood as 'silver of the Moon' (see VT46:18).
Before Mithril was reached there had been another intermediate form Thilevril (see below).
 

Names from The Treason of Isengard

Some of the manuscripts presented in this book were written on paper Tolkien had received in August 1940, others are still from the end of 1939 (TI:67). The story reached Lothlórien and the departure of Boromir by the winter of 1941-42 (TI:379,387) and continued then until Fangorn, the return of Gandalf and his visit to king Theoden in the same year, before taking a long break.

● Nen fimred (TI:10)
● Nenvithim 'Hoardales' (TI:114)
The region northeast of Imladris had been first named 'Dimrilldale(s)', but this name was then transferred to the east of the Misty Mountains, while this land received the names 'Hoardale(s)', 'Wolfdale', 'Entishdale', 'Entish Dales', 'Entish Lands' (with ent in the sense of 'giant' (RS:205)).
Both Elvish names seem to contain the plural of nan 'valley' (Let:230).
It is not entirely clear whether Nen fimred translates 'Hoardale' or 'Wolfdale'. The ending -red in fimred seems to denote an abstract noun, as in N. pathred 'fullness' (KWAT-), S. haered 'remote distance' (Rgeo:72, KHAYA-), but as the initial element fim 'slender' as in Fimbrethil 'slender-beech' (LotR App.F) or 'slender-princess' (√PHIM, PE17:23,82) does not fit at all and I have no sufficient explanation.
The second element in Nenvithim appears to be lenited #mithim *'hoar, grey' from MITH- with N. mith 'grey'. The adjectival ending -im is here very interesting. It must come from -imā, but would normally develop by lenition into -ef [ev] in Noldorin/Sindarin. No forms with -ef occur in these languages, however. Moreover, we are later explicitly told that the S equivalent of Q -ima (*-ef) was not current (WJ:387).
The absence of a-affection (triggering i > e) is not problematic here, conceptions were often changed and there was no such development in Ilkorin (compare also Celebrin and Palathrin above).
Both the absence of m-lenition and a-affection is known from the northern dialect of Sindarin, for instance lāmina > lomin instead of loven (PE17:133, VT41:10,  [6]). A similar conception could have been already present in these early writings. But another and probably easier explanation would be that this word is actually Ilkorin, which does not show m-lenition either, see e.g. dûm 'twilight' (DOMO-).
The same ending -im most probably occurs in Neleg Thilim 'the Gleaming tooth', see below.

● Orodnaur (TI:28,39)
The form Orodruin (UT:459) appears already here (TI:493), but once -naur was pencilled above -ruin. Orodruin 'Mountain of Blazing Fire' contains orod 'mountain' (ÓROT-) and, according to The Silmarillion index, S. ruin 'red flame' (Q. runya, compare RUN- 'red, glowing' (PM:366) and √RUYU 'blaze (red)' in PE17:183; but earlier it may have been either lenited gruin 'ruddy' (ROY2-) or #ruin as in Ruinnel (see below). As the second element the earlier form clearly contains N. naur 'fire' (NAR1-).

● Branduin >> Baranduin >> Malevarn (TI:66)
● Branduinen (TI:124)
The first two forms are listed under BARÁN-; #bran- *'brown, swart, dark-brown' obviously being a different contraction of the stem quite similar to #cran- from KARÁN- (see above). The second element is clearly duin 'river' (DUI-,VT45:11).
More interesting is the third form. It probably contains lenited #varn < #barn 'brown' with yet another different syncopation, also in analogy to #carn < KARÁN- (see above). The first element #male- must be related to SMAL- 'yellow', whence N. malt 'gold (as metal)', malthen 'of gold', malen 'yellow'. In the Nomeclature of the Lord of the Rings Tolkien translates Baranduin as 'the long gold-brown river' (RC:765) with S. baran 'brown, yellow-brown'. So it appears that with Malevarn he tried to form an actual compound *'gold-brown', but then decided that baran can already mean 'gold-brown', not just 'brown'.
The alternative form Branduinen should be probably analyzed as b'ran-dui-nen *'brown-flow-water' with dui- 'flow (in volume)' (RC:766) and nen 'water' (NEN-).

● Angrobel 'Irongarth' (TI:72) [Isengard]
Later on, the Sindarin name of 'Isengard' is Angrenost (LotRIII ch.4) with angren 'of iron' (ANGĀ-) and ost 'city, town with wall around' (OS-). This earlier compound contains lenited gobel 'walled house or village, 'town'' (PEL(ES)-) instead. The first element may be a differently formed adjective *ang'rā > N. *angr, *angor 'of iron'; here #angr- in the compound.
For another name see Angost below.

● Eärendillinwë 'The Short Lay of Eärendel' (TI:102-103) < Eärendil + #linwë '[short] lay', obviously from LIN2- 'sing'

● Annerchion 'Goblin Gate' (TI:114) - see Annerchin above and the note at the end.

● Ruinnel 'Redway' (TI:114)
● Celeb(rind)rath 'Silverlode' (TI:241)
These are further names of later Celebrant, earlier Crandir (and other forms, see above).
Although there is a later Sindarin word ruin '(fiery) red' (PM:366) < RUN- 'red, glowing', there is no hint that this stem was current at the time of The Etymologies. Instead we find RUN- as 'flat of hand and sole of foot', but ROY- (N. GROJ-) with gruin 'ruddy, red'. The explicit mention of Noldorin GROJ- instead of ROY- together with initial r- (which becomes rh- in Noldorin) lead to the assumption that this may be in fact an Ilkorin word from the same stem or from earlier RUY- (VT46:12), as Ilkorin keeps voiced r- [1]. Compare also later √RUYU 'blaze (red)' with derivatives like Q. ruina 'blazing, fiery'; Q. ruimen, S. ruist 'fireplace, hearth' (PE17:183).
However, in The Return of the Shadow Rimbedir, Rathcarn and Rohan were already found, on the other hand Rhascaron. Are the three first words then Ilkorin as well? It is hard to say, especially as there was no abrupt jump in the external development from Noldorin to Sindarin, which also keeps voiced r-.
If the translation 'Redway' is literal, the element -nel must inevitably mean 'way', but it is difficult to refer it to a stem. We find NEL- with the meaning 'three', originally 'point, triangle' (VT46:3). Perhaps we are rather dealing here with N. *nel(l), a cognate of Q. nelle (< nen-le) (NEN-), translated as 'brook', but here used for a river. See also Nen-uinel below.
Celebrath consists transparently out of celeb 'silver' (KYELEP-,TELEP-) and N. rath 'course, river-bed' (RAT-). The alternative Celebrindrath clearly contains the adjective celebrin 'of silver'. The intrusion of d into the cluster -nr- is probably mere epenthesis, compare G. in·runc > indrunc (GL:8) or *man- > mandra 'noble' (GL:56).

● Palath-ledin 'Gladden Field[s]' (TI:114)
For palath 'iris, gladden' see Palathrin above. The second element looks like the pluralized Noldorin adjective lhaden 'open, cleared' > pl. lhedin (LAT-). But this makes little sense, especially because palath is clearly singular. A better interpretation could be that the stem LAT- 'lie open' yields Ilk. *lad 'plain, field', which is then regularly pluralized in Ilkorin to *ledin, compare lad 'plain, valley' in Dagorlad, Himlad; imlad a narrow valley with steep sides, in Imladris (Silm.index). Noldorin sometimes also shows pl. -in, as nauglin 'dwarves' (NAUK-) for example, so we would deal with lenited *lhad > pl. *lhedin here.
Compare also Calledin below.

● Minas-tir, Minas-ond, Minas-berel (TI:115)
These are some earlier names of Minas Tirith.
All the forms contain minas 'tower' (minnas under MINI-). The first then appends the element tir 'watch, watch over' (Silm.index).
The second appends ond 'stone', probably still in the form without initial G-, see the commentary on Toll-onnui and related forms below (also GOND-, Silm.index).
The last form is more difficult to interpret. It could be related to BER- 'valiant' with ómataima and the ending -lā; *bere-lā > berel. This may be an adjective as sael 'wise' (SD:129) from SAY- or a noun, as N. cael, Q. kaila 'lying in bed, sickness' from KAY-; or both as N. (h)mael 'stain and adj. stained' < magla (SMAG-).
Hence the three words may be interpreted as: *'watching-tower', *'stony tower', *'valiant tower' or *'tower of valour'.

● Minos Giliath, Minos rhain, Othrain ' .... city', Minos tirith (TI:116)
This is another set of early names of Minas Tirith.
Minos is here probably a compound out of MINI- and N. ost 'city, town with wall around' (OS-), simplified -st > -s finally, as in Imladris < Imladrist.
The first form then appends a noun in genitival position - N. giliath, a collective plural 'stars' (GIL-) and it is interestingly unlenited, reflecting a change in the Noldorin grammar towards Sindarin. The same construction could be used in the second form with N. rhein, rhain 'border' (REG-), but we might be also dealing with the adjective rhain 'free' (RAN-, VT46:10). Both *'border city' and *'free city' (in the sense of 'free from the dominion of Sauron') make sense, but the presence of Emyn Rhain 'Border Hills' in the landscape rather suggests the variant *'border city'.
In Othrain medial -thr- seems to be the result of the contact between st and rh, i.e. < ost + rhain. Compare ost-rond > othrond 'fortress, city in underground caves' (OS-).

● Sirvinya 'New Sirion' (TI:119) < element sir- *'river' (SIR-) as in Sirion + Q. vinya 'new', cf. winya 'new, fresh, young' (VT45:16), 'young' (PE17:191) and later compounds as Vinyarie 'Newyear's Day' (PM:127)
● Andon (TI:308) *'long one' < N. and, ann 'long' (ÁNAD-, ANDA-) + name-formative or augmentative suffix -on (PE17:43, note)
These are earlier names of later Anduin, changed from Beleghir; the first one is unusually in Quenya.

● Ramathor, Ramathir >> Belfalas (TI:119)
These earlier names of the Bay of Belfalas are most difficult to interpret. It looks as if both are containing N. rhamb, rham 'wall' (RAMBĀ-) or a similar Ilkorin word with voiced intial r-.
The second element in the first form could be related to THUR- 'surround, fence, ward, hedge in, secrete' with N. thoro- 'fence'. The whole word thus could be translated as *'surrounding/fencing wall' - perhaps appropriate for a bay with high cliffs. Another stem in question is THOR-, THORON- with N. thôr, thoron, Ilk. thorn 'eagle'. And if the name is indeed Ilkorin, we may consider rama- as meaning 'wing' (RAM-) with unlenited m, thus ?'eagle-wing' - referring to the bay's shape on the map?
On the other hand, if one considers both early forms, it appears that Tolkien simply exchanged the ending -or for -ir (compare Brandor and Brandir below), which would leave us the basic form #ramath - interpretable as the collective plural *'walls', but perhaps also *'wings'.
But after all, we are later told that Bel- in Belfalas has no suitable meaning in Sindarin, probably meaning 'shore' in an alien tongue (VT42:15), so the earlier names need not to be entirely interpretable at all.

● Imlad-ril >> Imlad-rist (TI:120) < N. imlad 'dell, deep valley' (VT45:18) + -ril 'glitter, brilliance' (RIL-) or + rhis, rhess 'a ravine' (RIS-)
The first name of later Imladris seems to mean *'glittering/brilliant valley'. The following form is virtually identical to Imladris, which further simplifies -st to -s. But the writing with the hyphen is probably not done merely to indicate the parts of the word. Boromir's rhyme, following in the text, begins with the lines:
Seek for the Sword that was broken:
in Imlad-rist it dwells,
and there shall words be spoken
stronger than Morgol-spells. [...] (TI:128)
The metre requires a stress in Imlad-rist it dwells, so one can guess that Imlad-rist is meant to be stressed on the first and last syllable.

● Ilmandur *'Starlight-servant' (TI:119) [>> Isildur] < Q. Ilma 'starlight' (GIL-) + -(n)dur properly this means 'to serve', as one serves a legitimate master (Let:297)

● Ered Myrn 'Black Mountains' >> Eredvyrn (Mornvenniath) (TI:124)
These are earlier names of Ered Nimrais, the 'White Mountains' or 'White-horn Mountains' (see below).
The first two forms consist out of N. ered, pl. of orod 'mountain' (ÓROT-) and myrn, pl. of morn 'black' (MOR-,VT45:35), lenited to vyrn in the second word.
Mornvenniath has been analyzed by Patrick Wynne in Lambengolmor message #799 as probably containing N. ment 'point' (MET-), menn- medially and iâth 'fence' as in Doriath 'Land of the Fence' (WJ:370).

● Eredhithui, Hithdilias 'Misty Mountains' (TI:124)
These are earlier names of Hithaeglir, the 'Misty Mountains'.
The first form obviously contains N. ered, pl. of orod 'mountain' (ÓROT-) and hithui *'misty' (root KHITH- + adjectival ending -ui), not appearing under KHIS-, KHITH- 'mist, fog' in The Etymologies, but it appears as the Sindarin name of the month 'November' (Q. Hísimë) in LotR App. D.
The second word has N. hith 'mist, fog' and lenited #tilias, apparently a formation from TIL- 'point, horn', whence N. tild, till 'horn, referring to mountain-peaks'; also S. till 'tine, spike, point' (PE17:36) The usage of the ending -as (here with i-reduplication from the stem) may be compared with N. pân 'plank, fixed board, especially in a floor' in comparison with panas 'floor' (PAN-), thus #tilias - 'mountains' from a single 'peak, horn'.
If this word had been formed directly from tild, till, one would have expected *tildias, *tillias, however; so one should assume that it comes directly from the stem or by analogy from *til.

● Osforod >> Fornobel (TI:125,147)
This name appears later as Fornost or Fornost Erain 'Norbury of the Kings' (LotRVI ch.7); 'Norbury' from Old English north-burg 'north-(fortified) town' (RC:774).
The earlier names clearly mean the same, with slightly different elements. The former contains N. ost 'city, town with wall around' (OS-) and forod 'north' (PHOR-), the medial cluster -stf- being simplified to -sf-.
The latter contains forn 'right or north' (PHOR-, PE17:80) and lenited gobel 'walled house or village, 'town'' (PEL(ES)-).

● Gwaewar 'the Windlord' (TI:134) < N. gwaew (WĀ-,WAWA-,WAIWA-) + âr 'lord or king of a specified region' (TĀ-,TAƷ-, cf. ƷAR-)
Later this is the eagle Gwaihir. The name Gwaewar 'lord of the wind' already occurred earlier in LR:301 and was changed to Gwaihir as well.

● Tarkilmar >> Torfirion 'Westermanton' (TI:144-145)
These are earlier names of Annúminas, the capital of Arnor.
The former variant is clearly Quenya containing attested Tarkil (pl. Tarkildi) 'Great Men of Númenor', in other notes tarkil(di) 'high-men = Elf-friends of Númenor' (VT46:17, TĀ-,TAƷ-; KHIL-) and -mar 'home' (MBAR-, PE17:104-109), appearing in many place-names.
The latter variant seems to be a translation into Noldorin, torfir being the cognate of tarkil (ibid., PHIR-). The suffix -ion is then either name-formative with i-reduplication from the stem, adjectival or denoting a region, see the note at the end.

● Ered Orgoroth (Gorgoroth) 'Mountains of Terror' (TI:145) < N. ered, pl. of orod 'mountain' (ÓROT-) + lenited N. Gorgoroth 'deadly fear' (ÑGÓROTH-, ÑGUR-)

● Rhosgobel 'Brownhay' (TI:149,164,173) [dwelling of Radagast] < N. rhosc 'brown' (RUSKA-) + gobel 'walled house or village, 'town'' (PEL(ES)-)

● Celebras (Kelebras) 'Silverhorn' (TI:174) [later: Celebdil] < celeb 'silver' (KYELEP-,TELEP-) presumably + lenited ?rhass 'horn especially on living animal, but also applied to mountains') (RAS-, VT46:10), of uncertain reading in The Etymologies (S. rass 'horn' in PE17:36)

● Fanuiras 'Horn of Cloud' (TI:174) [later: Fanuidhol] < #fanui 'cloudy' (cf. Rgeo:74, SPAN-) + presumably lenited ?rhass 'horn especially on living animal, but also applied to mountains' (RAS-, VT46:10), of uncertain reading in The Etymologies (S. rass 'horn' in PE17:36)

● Thilevril (TI:184)
This was an earlier name of the substance Mithril, changed from Erceleb and Ithil (see above).
In The Etymologies we find Silevril (RIL-) as the Noldorin cogante of Q. Silmaril. Thilevril is obviously formed with the parallel stem THIL- instead of SIL-. Literally it would mean *'silvery brilliance'.

● Linglorel, Inglorel, Nimladel, Nimlorel (TI:223)
● Nimlothel >> Nimrodel (TI:223)
These earlier names of Nimrodel were analyzed in Lambengolmor message #927 by Philipp Marquart.
In short, Linglorel probably means something like *'woman of the golden tune' with LIN2- 'sing', N. lhind, lhinn 'air, tune' + glor- 'gold' in names (LÁWAR-, N. GLÁWAR-) + feminine suffix -el.
Inglorel may be *'woman of the golden heart' with N. ind, inn 'inner thought, meaning, heart' (ID-) + glor- 'gold' (compare Inglor < Indo-klār or Indo-glaurē under ID-).
Nimladel may be *'woman of the white valley' with N. nimp, nim 'pale' (NIK-W-), also S. nim 'white (usual word)' < Telerin nimbĭ, alteration of nimpĭ < *ninkwĭ < NIK- (PE17:49,168); + lad 'plain, valley' (Silm.index) (or maybe N. *lhad; compare Palath-ledin above).
Nimlorel is probably *'woman of the white gold' with nimp, nim and lenited glor-.
Nimlothel is probably *'woman of the white flower' with nimp, nim and N. lhoth 'flower' (LOT(H)-).
Finally, Nimrodel is translated as 'Lady of the Caves' (Silm.index) with groth, grod 'delving, underground dwelling' and as 'Lady of the White Grotto' (UT:457). The latter translation is more accurate, since the name obviously contains nimp, nim.
An earlier interpretation is that it contains nim, but the rest is unclear. If Sindarin it could be "lofty star," S. raud 'high, lofty, noble' < (a)rāta. But the possibility of S. -roth, rod < √ROT 'cave' is also mentioned. At the end Tolkien wrote: Or be a feminine, from rodel, lady, high lady.

● Ammalas >> Amaldor >> Amroth (TI:223)
By the time of 1968 Tolkien translates Amroth as 'upclimber, high-climber' (UT:245) because he had probably introduced the habit of living on telain (flats) up in the trees among the Galadhrim. Thus the initial element must be am- 'up' from AM2-, also √AM 'go up' (PE17:146). For the second element Christopher Tolkien cites rath- 'climb' (UT:255), whence also rath, applied in Númenórean Sindarin to longer road-ways and streets of Minas Tirith, as Rath Dínen 'the Silent Street'. In The Etymologies N. rath 'course, river-bed' is derived from RAT- 'walk'. However, according to a different conception Amroth and Nimrodel cannot be fully explained in Sindarin, though fitting it in form (UT:257). Similarly Haldir, cerin, caras, Orophin, Legolas, Thranduil, Amroth, Nimrodel, are suitable to Sindarin sounds and patterns, but are not clearly etymologizable as Sindarin (PE17:51). However, Nimrodel is translated several times (see above).
Anyway, it seems that the early names of Amroth also show the same prefix am- 'up'. In Ammalas it could be the extended variant amba- with -mb- > -mm- and the final element -las may be lenited N. lhass 'leaf' (LAS1-) or glas 'joy' as in the name Borlas (GÁLAS-).
In Amaldor the final element -aldor may be lenited galdor, itself appearing as a personal name from GALAD- 'tree'; the compound thus meaning something like *'person up on the tree'.

● Hathaldir (TI:227,240)
The Elf who meets the fellowship at Lórien was originally named Haldir; this was changed to Hathaldir and then back again to Haldir. 'Hathaldir the Young' also appears in the early Silmarillion as a companion of Barahir (LR:282). Later on, we encounter a Númenorean named Hatholdir (UT:173); and Hathol 'the Axe', a descendant of Marach (WJ:234).
In The Etymologies we find N. hathel < syatsēla, syatsĕla 'broadsword or axe-blade', formed with the ending -la denoting an object from syadsē > syatsē 'cleft, gash' (SYAD-). #Hathal would then require *syad-sa-la > *syatsala with a noun derivational ending -sā, -să instead of -sē, -sĕ. Such an ending occurs in the derivation of Q. apsa 'cooked food, meat' (AP-) for example. Hathol could be from *syatsāla with long ā > au > o.
The second element in the name is clearly N. †dîr 'male, man (elf, mortal, or of other speaking race)' surviving chiefly in proper names (DER-).

● Rhimbron (TI:227) >> Rhomrin, Romrin (TI:236)
● Rhimlath >> Rhimdir (TI:240)
These are the names of Haldir's brothers, the first one later becoming Rúmil and the second one Orfin, Orofin and eventually Orophin.
Rhimbron may contain the initial element #rhim- from the stem RIM- with N. rhemb, rhem 'frequent, numerous', rhimb, rhim 'crowd, host' and #-bron from BOR- BORÓN- with ON. boron 'steadfast, trusty man, faithful vassal' (> N. bór), brono- 'last, survive' and so on. Hence the name might be interpreted as *'he who endures a lot', perhaps in analogy to Rimbedir above. Remarkable is the appearance of the element her 'host, army' in Germanic names as Hermann, Walther; and rhim(b) 'crowd, host' might play a similar role here. Alternatively we should, as in the case of Rimbedir, consider RIP- 'rush, fly, flying', N. rhimp, thus *'rush-endure'. Many Germanic (and also Elvish) names are in fact put together similarly, out of two elements with or without an apparent connection to each other. But possibly -b- still belongs to the first element, while -ron is a name suffix; compare N. callon < kalrondō 'hero' (KAL-).
R(h)omrin looks like an adjective from N. rhom 'horn, trumpet', but -rin may also be a name-formative suffix.
For Rhimlath the best word in sight is N. lhath 'thong of [?leather]' (LATH-), perhaps in the sense of 'bow-string', thus *'rushing/fast bow-string'?
Rhimdir finally contains suffixed N. †dîr 'male, man (elf, mortal, or of other speaking race)' surviving chiefly in proper names (DER-), so it might be *'the rushing one'.
However, these are just speculative suggestions. It is very well possible that these names do not bear a significant meaning at all, or at least not all of them.

● Naith 'Angle' (TI:268, LotRII, ch.6)
● Nelennas >> Nelen 'the Gore' >> Narthas (TI:231,236,242)
● Bennas (TI:238,241,288)
● Egladil (TI:288, LotRII, ch.8)
All these words refer to the triangular shape of Lórien at the map pointing as a cape into the Anduin and are translated as 'gore' or 'angle'.
Naith appears under SNAS-, SNAT- as 'gore'. Also S. neith, naith, Q. nehte 'angle' < nek-tē < √NEK (PE17:55).
Nelennas could contain #nelen *'triangular' from NEL- 'three' with adjectival -en (cf. S. neg(e)n 'sharp, angular' < √NEK) and N. nass 'point, sharp end; angle or corner' (NAS-). The stem NEL- was originally glossed 'point, triangle' (VT46:3), which would explain the sole Nelen *'triangular point' - now with the noun ending -en (see Lamben directly below) and without -nas.
Narthas 'gore' had been written under NARTA- 'spear point, gore triangle' (VT45:37) before the meaning of this stem was changed to 'kindle'.
Also bennas 'gore' is directly attested, given under BEN- 'corner (from inside), angle'.
Egladil found a way into The Lord of the Rings and is translated as 'elven-point' in the index (RC:307). The final element is clearly lenited #-til 'point, horn' (TIL-), S. till 'tine, spike, point' (PE17:36). Egla is attested as the Doriathrin word for 'Elf', also in compounds like Eglamar or Eglorest (ELED-) and was thus maybe originally intended to be dialectical. Later we find Eglan, pl. Eglain, Egladhrim 'The name that the Sindar gave to themselves' ('the Forsaken')' from HEKE 'aside, apart, separate'; and shorter hekla or heklā surviving in place-names like Eglamar (WJ:365) - these words are now pure Sindarin of course.
It is interesting that #egla- may be itself be interpreted as 'point, gore', related to EK-, EKTE- 'spear' with N. êg 'thorn', derived with -la as makla 'sword' form MAK- (> N. magl, magol). But this this is probably just coincidence.

● Calendil 'Green Spit, Green-tine', {Nelen} Calennel (TI:268,288)
● {Lamben 'Tongue'} (TI:280)
This is another set of names for the triangle of Lórien.
Calendil transparently consists out of calen 'bright-coloured = green' (KAL-) and suffixed #-til 'point, horn' (TIL-), S. till 'tine, spike, point' (PE17:36); while Calennel has suffixed -nel, probably from earlier NEL- with the meaning 'point, triangle' (VT46:3).
Also very interesting is the variant Lamben 'Tongue' (see LAB- 'lick', N. lham(b) 'tongue'). Medial -mb- is possible within the phonology of Noldorin, also if both phonemes belong to the same root, compare for instance ambar, amar 'Earth' (MBAR-). Although the ending -en dominates in Noldorin/Sindarin adjectives, it is also found in the derivation of nouns, e.g. N. lhalwen 'elm-tree' (LÁLAM-), S. Ceven 'Earth' (Q. kemen) (VT44:21, KEM-) and also Nelen 'the Gore' above.
This may lead to a surprising result - in The Lord of the Rings we see Gandalf saying lasto beth lammen 'listen to the word of my tongue' (LotRII ch.4, RS:463). Since lamben is directly attested as 'tongue', one should consider the possibility of a conception that he does not use any possessive suffix -en 'my' at all. PE17:46 clearly cites lammen '(of) my tongue', but on another note the adjective lammen, lambina 'of tongue, spoken with tongue'.
A different possibility is that Lamben is a compound of lam 'tongue' and BEN- 'corner (from inside), angle' (cf. Bennas above), but if so, it is not indicated by the simple translation 'tongue'.

● Dol Dúgol >> Dol Dúghul (TI:234,296) < dôl 'head' (NDOL-) or dol 'head, hill' (RC:433), 'head; often applied to hills and mountains' (Silm.index), S. dol (doll) (PE17:36) and N. dû 'nightfall, late evening' or 'night, dread of night' (DOƷ-, DÔ-) + gûl 'magic' (ÑGOL-); lenited to -ghul in the latter case
This is the name of Dol Guldur, a hill in Southern Mirkwood. The compound of dû and -gul or -gol results in 'dark magic, sorcery'. Tolkien was not sure about the vowel here, compare dûghol, morgol >> durgul, mor(n)gul 'sorcery' (VT46:3). Long ō would result in long ū, shortened in the compound, while short o would remain.
Morgol also appears in Minas Morgol (TI:120).

● Bair am Yru ‘the houses of the Galadrim’ (TI:243) < bair, pl. of N. bár 'home' (VT45:33, PE17:104-109) + N. am 'up' (AM2-) + **yru, a very probable misreading for *yrn, pl. of orn 'tree, high isolated tree' (ÓR-NI-); thus literally: *'houses/homes up in the trees'

● Finduilas >> Rhien >> Galdrin (perhaps miswritten for Galdrien) (TI:249,262)
● Galadhrien = Galað-rhien 'tree-lady', Galadrien (TI:249)
These are earlier names of Galadriel.
Finduilas occurs later in The Silmarillion as the daughter of Orodreth (she fell in love with Túrin and was captured during the sack of Nargothrond). According to the index her name contains fin- 'hair'; findui looks like an associated adjective; and -las may be either 'leaf' (LAS1-) or 'joy' (lenited glas as in Borlas (GALÁS-)). On the other hand we may be dealing with the lenited element #-tui- 'spring' (TUY-), #tuilas 'spring leaf', hence maybe *'hair like leaves in spring'?
The next form Rhien is most probably identical to rhiend 'queen, lit. 'crowned' or crown-lady' < rīʒende, given in a deleted note under RIG- (VT46:11).
Then, Galdrien seems to mean 'tree-queen', from a syncopated #gald- (see GÁLAD-) with suffixed -rhien 'queen'.
The last form is quite self-explanatory with the historical form, now containing the full form galadh 'tree' instead of #gald-.

● Tar >> Aran >> Galdaran (TI:261,262)
● Galathir = Galað-hîr 'tree-lord' (TI:249)
● Arafain (TI:256)
These are earlier names of Celeborn.
The first two forms must simply mean 'king' - the former is most probably from the mentioned TAR- (VT45:17) or TĀ-, TAƷ- 'high, lofty, noble' (compare also Tarkilmar), the latter is given under ƷAR-; thus the Lords of Lothlórien are Aran and Rhien 'King' and 'Queen'.
Galdaran contains #gald-, obviously a contraction of GÁLAD- + aran 'king', thus meaning *'tree-king'. Hence, the Lords are now Galdaran and Galdrien *'Tree-king' and *'Tree-queen'. Although d becomes lenited in Noldorin after a vowel, as in galadh 'tree', the medial combination -ld- is apparently allowed - cf. Galdor, later Gallor (GALA-, GÁLAD-).
Galathir is again self-explanatory with the indicated etymology, formed with galadh 'tree' (GÁLAD-) and hîr 'master' (KHER-).
And finally the last form should probably mean *'white king' with the element ara- 'high, noble, royal' (Silm.index), also S. ara *'king', ar/ara/aran 'prefixes of excellence especially in royal names' (PE17:147) and fain *'white' (The Etymologies give just fein 'white' under SPAN-, but compare for instance the variation in N. lhein, lhain 'free(d)' (LEK-)).

● Ogodrûth 'Entwash' (TI:250)
This is the name of later Onodló (UT:459) or Onodiōl (RC:334) < onod 'ent' + lô, #iôl 'fen(land)'.
Here, #ogod 'ent' is probably still in the Old English meaning 'giant' (see RS:205); but and the Elvish name is of unclear origin, just as later onod. In The Etymologies we find deleted KHAN-AK-, Q. hanako 'giant' (VT45:21) and NOROTH-, Q. norsa 'giant', *N. Noroth, all being completely different.
But also #rûth *'flood-water, fenland' is difficult to relate to any other root or word. Adunaic #rôth 'foam' < Rôthinzil 'Flower of the Foam' (SD:360) comes somewhat close, while S. rûth 'anger' or 'wrath' (Silm.index, PE17:183,188) does not fit here at all.

● Toll-onnui >> Toll Ondren 'Great Carrock' (TI:268)
● Toll-ondu >> Toll-onnui >> Toll-ondren 'the Great Carrock' (TI:268)
● Tol Galen (TI:271)
● Tolondren >> Eregon >> Brandor >> Tol Brandor (TI:285)
● Tolharn >> Tollernen 'Stoneait' >> Eregon 'Stone pinnacle' (TI:324,345)
● Emris >> Eregon >> Tolbrandir (TI:367,316-318)
These are earlier names of Tol Brandir, a little island in the Anduin (cf. the drawing in PE17:22).
'Carrock' is related to Old English carr, Welsh carreg 'rock, stone' (RC:207) and most of the early names clearly mean *'stony island' with N. toll 'island' (TOL2-) and different adjectives *'stony', derived with the endings -ui, -ren; and -u < -wā/-wē in ondu (cf. N. cadu 'shaped, formed' < ON katwe/katwa (KAT-, VT45:19)).
There is the stem GOND- 'stone' in The Etymologies, so that one may think that all these forms are lenited *gonnui, *gondren, *gondu; but there was no initial G- originally - the first names of Gondor were Ond >> Ondor (TI:493), changed to Gondor later in the process of writing (Feb.9 1942, TI:423). In devising this word for Tolkien was influenced by a book from his childhood, where ond 'stone' was given as being known from a pre-Celtic language of Britain (Let:324). Note however that gonn 'great stone, rock' (GL:41) already appears very early in the Gnomish Lexicon and is associated with the name Gondolin.
'Galen?' with a question mark is once written above ondren, thus 'Green Isle' (with lenited N. calen 'bright-coloured = green' (KAL-)) was considered (cf. LR:268, Silm.index).
Eregon may mean *'single stone'. The adjective ereg occurs later as a name of a river in Ond(or) with the translation 'first' (see below). The element er- was always present in Tolkien's writings, sometimes used in counting series as 'one' or 'first', but with the original meaning 'remain alone' (QL:36) or 'be alone, deprived' (ERE-).
However, the translation 'Stone pinnacle' suggests rather ereg *'pinnacle'. The Etymologies give N. ereg 'holly-tree' from ERÉK- 'thorn', here perhaps in a more general meaning. The final element of Eregon should then be simplified ond 'stone'.
Then the name reaches a form close to later Brandir. Tolkien writes in the manuscript of Nomenclature that Brandir was of uncertain origin and meaning; prob[ably] a corruption of *baradnir Grey-elven (Sindarin) for tower-steep = "steep tower" (RC:333). Of course, barad 'tower' is the same as in Barad-dûr, 'The Dark Tower' (UT:422, BARAT-), but #nir 'steep' is difficult to relate to anything else and seems to be an ad-hoc invention.
Another interpretation is that Brandir is the plural of brand 'steeple' with the suffix -ir (cf. Thinnir *'Grey-elves' < thind (PE17:140)), hence Tol Brandir 'Isle of the Great Steeples' (PE17:61,22).
Brandor could perhaps be a compound of brand, brann 'lofty, noble, fine' (BARÁD-) or 'towering; tall and massive' (PE17:61) and the element #-or from ORO- 'up, rise, high', here indicating a high rock and used instead of the full noun orod 'mountain'. Compare Erebor 'The Lonely Mountain' from The Hobbit, which similarly contains N. ereb 'isolated' (ERE-) and #-or.
But of course -or may be just a name-formative suffix, as in the personal name Galadhor, Galdor from GALAD-, galadh 'tree'.
Tolharn obviously contains lenited sarn 'stone as a material; or as adj.' (SAR-) and Tollernen seems to include #ernen *'single' < *erninā, an adjective from ERE-.
Emris, at last, is difficult to explain. N. rhis, rhess 'a ravine' (RIS-) as in Imladris does not seem to fit in reference to an island, unless the cutting of a river is meant. This name is probably of an intendedly obscure origin.

● {Duil} Emyn Rain >> Emyn Rhain 'Border Hills' (TI:268,287,314) < N. emyn, pl. of amon 'hill' (AM2-) + N. rhein, rhain 'border' (REG-)
This is the name for a hill range northwest of Mordor (see the map on page 314).
In the former example rhain looks to be lenited according to the rules of late Noldorin (e.g. Ennyn Ðurin Aran Vória 'The Doors of Durin Lord of Moria' (TI:182)), in the latter it is not as in subsequent Sindarin (e.g. Ennyn Durin Aran Moria (LotRII ch.4, Rgeo)).
The struck through Duil 'hills' also occurs in Duil Rewinion (see below).

● Palath Nenui (TI:268,287,308) 'Wetwang' < palath 'surface' (PAL-), N. nen 'water', adj. nenui *'watery' not otherwise attested
This is a name of later Nindalf 'Wetwang'(RC:334).

● Rhosfein >> Dant-ruin, Dant-ruinel (TI:273,283,285)
These are earlier names of the waterfall Rauros 'Rush-rain, Roar-rain' (TI:285).
Rhosfein also appears as Rosfein (TI:315). It contains N. rhoss 'rain', also occurring in the waterfall name Celebros 'Silver-rain' (ROS2-), and fein 'white' (SPAN-), thus meaning *'white rain'.
The other two forms contain dant 'fall', cf. DAT-, DANT- 'fall down', N. dant- 'to fall' (later: MR:373).
The elements ruin and ruinel are more difficult in this context. Although Ruinnel 'Redway' occurred earlier, there is no hint that this should be a *'red fall', but without a translation this cannot be said with full confidence, of course. Judging by 'Rush-rain, Roar-rain', ruin might be somehow related to N. †rhû 'sound of horns' (Q. róma 'loud-sound, trumpet-sound', ROM- 'loud noise'). Another possibility might be a derivation from RŌ- 'rise', yielding words for 'east', thus perhaps *rōnya > N. *r(h)uin 'eastern'? It might be also possible that ruin is just as later #rau- derived from RAW-, which yields words for 'lion', but might have been an original onomatopoetic imitation of a roaring sound. In either case one would require an ending *-njā to arrive at the medial diphthong ui, compare N. rhaw 'lion', pl. rhui < rōvi. Evidence for such an ending may be the formation of Q. menya 'our' < *me-njā (VT43:19). Compare also echain (presumably from KAT-) below.
In ruinel we observe the suffix -el. It often appears in personal feminine names, but also in such noun formations as N. gandel, gannel 'a harp' (ÑGAN-, ÑGÁNAD-), findel, finnel '(braided) hair' < sphíndele (SPIN-) or nelladel 'ringing of bells' (NYEL-), perhaps also sarnel *'statue' (see below). This may confirm the assumption that ruin does not refer here to the colour 'red' (as *'fall of redness' would be most unusual). If the guess with the loud sound is right, the noun 'roaring' might have been formed here.
Alternatively we could have in the last example #rui- *'loud, roaring' (< *rōmjā?, compare also brui 'loud' < Bruinen 'Loudwater' (RC:171), G. rum, brum 'noise' (GL:66)) + #nel, cognate of Q. nelle (< nen-le) 'brook' (NEN-) as possibly in Ruinnel.
Ruin also occurs in Sarn Ruin, see directly below: Compare also Nen-uinel.

● Sarn-Ruin, Sarn Ruin (TI:318,320)
● Pensarn (TI:353,360)
These are earlier names of Anduin's rapids, called Sarn Gebir in The Fellowship of the Ring.
All names include sarn 'stone as a material; or as adj.' (SAR-). For ruin (here possibly *'rushing, roaring') see Rhosfein above.
Pensarn is perhaps derived from PEN-, PÉNED-, whence N. pend, penn 'declivity', Q. penda 'sloping down, inclined'; also penn 'slope' (RC:525) + sarn; thus *'stony slope'.

● Eldamir, Eledon 'Elf-stone', Qendemir (TI:276,366)
● Eldakar 'Elfhelm' (TI:276,366)
● Eldavel 'Elfwold' (TI:366)
The first line brings up another set of of Aragorn's (Trotter's) early names. Eldakar and Eldavel were considered as the names of his father.
Eldamir is clearly Quenya, consisting out of Elda 'Elf' (ELED-) and #-mir 'jewel' (MIR-, Silm.index). Eledon would then be the Noldorin cognate, containing Eledh 'Elf' (ELED-) and -ond 'stone', not lenited gond, gonn (GOND-, Silm.index), but rather still in the basic form ond, *onn - see the commentary to Toll-onnui and related forms above. Tolkien often spelled dh as d (as also e.g. Caras Galadon = Caras Galaðon) because according to his own words dh is not used in English and looks uncouth (UT:267).
Qendemir is then again Quenya with qende 'elf' (KWEN(ED)-) and #-mir 'jewel'.
For final -kar in Eldakar compare #karma 'helm' from Karma-kundo 'Helm-guardian' (PM:260), Q. kár 'head', kassa 'helmet' (KAS- 'head') and also the explanation that -kar is from kāsā < √KAS 'head' (PE17:114).
The name Elfwold is akin to Old Norse Alfvaldr meaning 'elf-ruler' and the Quenya name has suffixed #-vel, most probably from BEL- 'strong' with Ilk. bel 'strength' and similar derivatives.

● Uvanwaith 'Nomenlands', 'Noman's Land' (TI:281,283,314)
This is the land lying between the Border Hills (Emyn Rhain, see above) and Mordor.
The name has been discussed in Lambengolmor #847 (and answers). While the final element is clearly lenited gwaith, literally 'manhood, also used = man-power, troop of able-bodied men, host, regiment' (weith under WEG-, VT46:21) and is by extension used to denote a region in many place-names as in Enedwaith 'Middlemarch' (TI:304), there seem to be two possibilities for the etymology of the first element.
It could be formed by the negative prefix #uv- (lenited from UMU-, see UGU-) + N. anw 'male' (ƷAN-), compare Q. Úner 'Noman' (UT:211), which must be formed with nér 'man' (NDER-, MR:213) or Goldogrin unweg 'nobody, no one' (GL:73), formed with gweg 'man, male of Elda or Indi' (GL:44).
Another possibility may be a formation with negative u- as in úan 'monster' < ūbanō (BAN-) and lenited #man, a general pronoun *'somebody, someone'. It is only attested as an interrogative, e.g. in Namárië; but compare the indefinite suffixed article of Early Qenya -ma (PE14:42) and the later pronoun ma 'something, a thing' (VT42:34), which may be also contained in the interrogative mana?, compounded *'what-is?' (PM:395). See also mo 'somebody, someone' (VT42:34).

[● Duil Rewinion 'Hills of the Hunters' (TI:287)
This place later appears as Taur-na-Faroth (LR:424, SPAR-) and is not part of the drafts for The Lord of the Rings, but I have included it nevertheless.
The first element must be the plural of dôl 'head' (NDOL-) or dol 'head, hill' (RC:433), 'head; often applied to hills and mountains' (Silm.index), S. dol (doll) (PE17:36) by simple i-affection with o > u. However, Noldorin affects o > y and ô > ŷ in monosyllables [3] in The Etymologies. Ilkorin usually uses the plural suffix -in, but in one case at least a monosyllabic noun is pluralized by i-affection - Ilk. tal 'foot', pl. tel (TAL-), so that the possibility of an Ilkorin word remains. In the name Emyn Rain (see above) Emyn overwrites the struck through Duil.
In the second word the ending -ion [note] is attached to a basic form, which can be deduced as #rawen or #rewin. This is most probably formed from RAW- yielding Q. rá, N. rhaw 'lion' (compare RAB- yielding Q. ráva, N. rhaw 'wilderness').
Seen in a larger context a derivation of 'hunter' form such a root is not new, compare Gnomish rau 'lion', raust 'hunt, chase', rautha- 'hunt, chase, pursue' (GL:65), also Qenya roa 'a wild beast', rauste 'hunting, preying', rauta- 'chase, hunt, pursue - extirpate, exterminate', rawa- 'run, chase' and so on; stems RAVA, RAWA (QL:79).
The formation of 'hunter' may be done here via the suffix -in, a cognate to Q. -indo, which can be for instance isolated from melindo 'lover' (MEL-), thus: *raw-indō > *rewin(n). For this perhaps compare melin (mellyn, meldir) above.]

● Rhimdad, Rhimdath, Rhibdath 'Rushdown' (TI:296)
Compare Rhimdath above and RIP- in The Etymologies - the first form does not appear there, but is apparently formed with dad < *dat- 'down' (DAT-, DANT-) instead of dath < dattā.

● Forlorn 'North Haven' (TI.301-302)
● Harlorn 'South Haven' (TI.301-302)
These are two havens to the north and south of the Gulf of Lune.
The initial elements are clearly #for- 'north' (PHOR- 'right-hand') and #har- 'south' (KHYAR- 'left hand') respectively. A later note dated Feb. 9 1942 reads: lorn 'haven' and londe 'gulf' (TI:423). In The Etymologies, Q. londe 'road (in sea), fairway, entrance to harbour' and N. lhonn 'narrow pass, strait, pass' (later 'haven', e.g. Lond Daer 'Great Haven' (UT:450)) are both derived from LOD-, but there is also a direct gloss of N. lhorn 'quiet water, anchorage, haven, harbour' from LUR- 'be quiet, still, calm' (VT45:29).
This is a good example of the external development from Noldorin with L- yielding voiceless lh- (lhorn) towards Sindarin, which keeps voiced l- (lorn).

Now follow earlier names of the rivers of Gondor. Their final forms, which were ultimately published, are explained in VT42:7-18. As it can be seen from these late explanations, the linguistic situation in that area was very complicated and not all names of the rivers are even Sindarin. Most of them already occur in these early writings:

● Ithilduin, Duin Morghul (TI:312)
The former name clearly means *'moon-river', consisting out of N. Ithil 'poetic name of the Moon' (THIL-) and duin 'river' (DUI-, VT45:11).
The latter must mean *'river of dark magic', the second word consisting out of the element mor- 'black', appearing in many other names (see MOR-) and lenited gûl 'magic' (ÑGOL-).

● Poros 'Boundary', Ereg 'First', Sirith (TI:312)
On Sirith see VT42:11.
Ereg appears later as Erui (VT42:10), the formation of the earlier form seems to be done with the adjectival suffix *-kā > -g; compare SPAY- > Q. faika 'contemptible, mean', N. foeg 'mean, poor, bad' and deleted tadeg 'second' (VT42:10) < TATA-, ATTA- (VT42:24).
Poros is difficult to interpret; it would require primitive *por-, *kwor- or *pār-, *kwār- but POR- with Q. pore 'flour, meal', KWAR- 'clutching hand' (ON pōre 'fist'), PAR- 'compose, put together' do not seem suitable; so it is probably one of the not translatable names and 'boundary' may refer to the river's role, not its meaning.

● Lameduin (of Lamedon) with tributaries Serni, Kelos (TI:312)
For these see VT42:11,17. Later, Lameduin does not occur; but it appears to be partly or totally interpretable, containing duin 'river' (DUI-, VT45:11) and perhaps some derivative of LAM-, which produces words for 'sound' and 'echo' in The Etymologies, hence Lameduin < *lamme-duin *'echoing river'.

● Ringlo, Kiril, Morthond, Calenhir (TI:312)
See VT42:13-14. Calenhir is obviously the *'bright river' or *'green river' < calen 'bright-coloured = green' (KAL-) + lenited sîr 'river' (SIR-).

● Lhefneg 'Fifth' (TI:312), Odotheg, Odothui 'Seventh' (TI:312)
Lhefneg appears later as Lefnui, Levnui (VT42:14) with a different adjectival ending. Here, Lhefneg and Odotheg both show *-kā > -g as Ereg above. The stems are LEP- (LEPEN-, LEPEK-) 'five' with N. lheben ' and OT- (OTOS-, OTOK-) 'seven' with N. odog. For odothui 'seventh' see VT42:25.

● Barangils 'Swertings' (TI:309,313)
This is a Gondorian term for the people of Harad; 'swerting' derives from swart ('swarthy') (RC:764). Later Harns is also met (see below).
The Elvish name seems to contain baran 'dark, swart, dark-brown' (BARÁN) and maybe the suffix -il also seen in ernil 'prince' (LotRVI ch.4, UT:428) < S. #arn 'royal' (Let:347). If so, the occurrence of -g- looks unusual, but might be compared with S. fing 'lock of hair' (RC:386, SPIN-). In this example, however, -g- is perhaps an epenthetic insertion by speakers of the Common Speech; at least the name is given the English (= Common Speech) plural -s.
However, Matthew Dinse made me aware of another possibility. In 1932 and 1934 Tolkien published the two parts of an article called Sigelwara Land, which dealt with the question of why there was a distinct and several times attested Old English name for the Ethiopians (namely Sigelwaran, Sigelhearwan). This fact is special because proper biblical names were usually adapted, not translated.
Tolkien remarks that the word includes elements not current in Old English and argues from there on that it must be older and preserved at least a name, if no more, from the vanished native mythology or its borderland of half-mythical geography.
He then attempts a linguisitc analysis. As the initial element Sigel 'Sun' comes into question, which is attested in Beowulf (and other sources) and has furthermore cognates in other languages, as e.g. the name of the s-rune. Another candidate is sigel, sigle describing a round jewel or golden ornament, originating from Latin sigillum, which in its turn refers to a small image or figure, the impress of a stamp or seal. Tolkien then suggests that the two words had mutual influence on each other; and he remarks on the usage of gimm 'precious stone, jewel' for the sun.
For the second element hearwa he discusses several primitive candidates all having to do with the colour 'black', so that the name may mean something like *'those who were made black by the sun'.
As he concludes, such guesswork is naturally inconclusive, but not pointless - giving insight into English and northern trandition and imagination.
Now the interesting point is that Tolkien considered Harwan, Silharrows, Harrowland, Sunharrowland as names for Haradwaith, the very region occupied by the Swertings (TI:435,439). So it stands to reason that Barangil could mirror the formation of Sigelhearwan, containing baran 'dark, swart, dark-brown' as mentioned above and as the second element #-gil 'star', orginally 'bright spark' (N. geil under GIL-,VT45:15; later sg. gil (Rgeo:73)); but now with reference to the sun (= bright spark, jewel(?)). Although gil (or N. geil, Early Noldorin gail (PE13:143)) is always attested in the meaning 'star' and never has to do anything with the sun; one could imagine a corruption or alteration of this sense by the Gondorians, especially regarding the obscure Sigelwara. Another possibility would be N. bara 'fiery, eager', originally 'hot, burning' (BARAS-) and gil lenited to -ngil. Although the root is given as GIL- in The Etymologies, it later becomes √NGIL 'silver glint' (MR:388, PE167). Note that Tolkien considers a connection between -hearwan and Old Norse hyrr 'fire', Old English heorþ 'hearth', hierstan.'roast', so that an ambiguity of the Elvish word could have been intended.
Compare also Narghil, Nargil directly below.

● Narghil Pass >> Nargil Pass (TI:309-310)
This is the name of a mountain pass from Lothlann to southern Mordor; later it remains as Nargil.
As Matthew Dinse pointed out to me, Nargil likely contains lenited N. cîl 'cleft, pass between hills, gorge' (KIL-). Two possibilites remain for the initial element - either syncopated #narg- from NÁRAK- 'tear, rend (tr[ansitive] and intr[ansitive])' (compare Nargos below) or nar- from NAR-1 'flame, fire', perhaps referring to the colour 'red' as Narosîr and Narodûm above or Nargalad, Nardol below. We would gain possible tentative translations *'Rending Gorge', *'Red Pass' or *'Fire-Pass' (the latter perhaps referring to heat or the hostile environment of Mordor).
Earlier Narghil may be explained by the fact that KIL- 'divide' has, according to its entry a parallel stem SKEL-, although it does not appear anywhere else in The Etymologies. But as it is evident from other SK-stems, this initial cluster yields h- in Noldorin. Therefore, Narghil could contain unlenited *hil, *hîl 'gorge, pass' and #narg- would obviously remain as the initial element.
Finally, there is still at least a slight possibility that both forms contain gil or lenited ghil with reference to the sun, see Barangils above.

● Anarion (TI:309-310) < N. Anar 'Sun' (NAR-1) beside Anor < ANÁR- + -ion 'region' (RC:772, [note])
This is the name for the region western of the Anduin, later also Anárion (WR:243-244) with a long vowel or maybe stress indication; Anórien in The Lord of the Rings.

● Dagras (TI:310,389) < N. dagr, dagor 'battle' (NDAK-) + collective suffix -as, cf. panas 'floor' compared with N. pân 'plank, fixed board, especially in a floor' (PAN-) or #tilias 'mountains' from TIL- 'point, horn' in Hithdilias above
This is an earlier name of Dagorlad 'Battle Plain' (UT:427).

● Alfobas (TI:322) 'Alqualonde' <  N. alf 'swan' (ÁLAK-) + -hobas 'harbourage' (KHOP-)
This name is also directly given under KHOP-.

● Eredwethion ‘Mountains of Shadow’ (TI:325,345) < N. ered, pl. of orod 'mountain' (ÓROT-) + lenited and umlauted N. gwath 'shade' (WATH-) + ending -ion, for which see note
This name is already given under WATH- (explicitly Noldorin), the Ilkorin cogante being Urthin Gwethion (VT46:21).

● Amarthon, Dolamarth 'Mount Doom' (TI:343)
Both forms include ammarth 'fate, doom' (MBARAT-). According to LotR App.D mb became m in all cases, but still counted as a long consonant for purposes of stress [...], and is thus written mm in cases where otherwise the stress might be in doubt. The word amarth may only be stressed on the first syllable, so stress is not in doubt here and it can be safely written like that. Amarthon is stressed on the second syllable, so mm is not important here either. But one wonders why Dolamarth is written this way, since m < mm < mb would attract the stress to the penultimate syllable and one would expect *dolámmarth, not *dólamarth.
Anyway, the first form shows the name-formative or augmentative suffix -on (PE17:43), compare Andon above.
The second form (written above the first) has initially dôl 'head' (NDOL-) or dol 'head, hill' (RC:433), 'head; often applied to hills and mountains' (Silm.index), S. dol (doll) (PE17:36); and amarth in genitival position.

● Gorgos 'Dire-castle' (TI:344)
● Nargos (TI:344)
These are the names for two guarding towers of Cirith Ungol, later Nelig Myrn, Naglath Morn (see below). In The Lord of the Rings they are Narchost and Carchost, 'Towers of the Teeth' (LotRVI, ch.1).
In the first name, #gorg- 'dire' must be related to ÑGOROTH- 'horror', also gorgor 'extreme horror' (WJ:415) + -os, apparently a simplified N. ost 'city, town with wall around' (OS-), here - 'castle'.
In the preceding passage the two towers are called 'Loath and Grim' >> 'Fell and Dire', so that Nargos likely means *'Fell-castle'. If so, #narg- is probably syncopated NÁRAK- 'tear, rend (tr. and intr.)' with primitive narāka 'rushing, rapid, violent' > Q. naraka 'harsh, rending, violent', N. Narog as a river name. This stem could likely produce 'fell' in the sense of 'fierce, destructive'. Compare also narch 'bitter-biting' in later Narchost (RC:601).
But since the translation 'Fell and Dire' does not contain 'castle' the second part might also be lenited gost 'dread, terror' (GOS-, GOTH-) rather than ost, compare N. Tauros 'Forest-Dread'.

● Nen-uinel (TI:364,361), Kerin-muil (TI:364)
These are earlier names of Nen Hithoel 'Mist-cool Water' (UT:434), a lake formed by the river Anduin before Rauros.
The first form contains N. nen 'water' (NEN-), but the second element is difficult to interpret. It may contain ui- 'eternal' as in N. Uilos 'everlasting snow' (OY-) or it may be related to UY- with Q. uile 'long trailing plant, especially seeweed', [N.] uil 'seaweed', but then giving #-nel a suitable explanation is difficult. The river Ruinnel 'Redway' occurred earlier and I supposed that #-nel might be there the cognate of Q. nelle (< nen-le) 'brook' (NEN-); perhaps it is used here with a changed meaning for a pool or lake?
But we may also analyze the name as uin-el and would then deal with the same ending as in Dant-ruin, Dant-ruinel above. In Goldogrin at least, uin was associated with water, with the archaic meaning 'wave' and it was also the name of Ulmo's whale (GL:74). A common word for 'whale' was uimoth 'sheep of the waves' with moth 'sheep' (GL:58). However, in Early Noldorin uin (< *win-) means 'woman' (PE13:155).
The second form seems to contain N. cerin 'circular enclosure' (KOR-), later 'mound' (UT:426), 'circular mound or artificial hill' (TI:242-243) - but here perhaps just describing a place enclosed by a circular formation of hills. Muil also occurs in the name Emyn Muil, which is translated as 'Drear Hills' in the index (UT:434), so that it would mean 'drear'. But originally it may have been intended to mean 'hidden', from the stem MUY- with the adjective ending -l < -la as e.g. in S. fael 'fair-minded, just, generous' (Q. faila, PHAYA-; PM:352). Although a note reads not in Noldorin because it became identical with moina, this stem had earlier produced N. †muin 'secret' (VT45:36). Hence, Kerin-muil might be *'hidden enclosure'. In the Fellowship of the Ring Nen Hithoel is fenced by steep grey hills (LotRII ch.9).
Another explanation would be G. muil 'tarn' cognate to Q. moile (GL:58) already appearing in The Lost Tales contained in Umboth-muilin. Then kerin would be an adjective *'round, circular' (cf. Ramas Coren below) and hence Kerin-muil *'round tarn/lake'.

● Sern Aranath (TI:366) < sern pl. of sarn 'stone as a material; or as adj.' (SAR-) + aranath, class-plural of N. aran 'king' (ƷAR-), thus *'stones of the kings'
This is an earlier name of Argonath < arn(a)gon-ath 'the group of (two) noble stones' (Let:347), later also Sern Erain and other related forms are met (see below).

● Llawhen, {Amon Tirlaw, Lhawdir, Lasthen, Henlas}, Hendlas (TI:387)
● Tirmindon, Larmindon (TI:364)
The set of names in the first line is an agglomeration of experimental forms for what should later become Amon Lhaw and Amon Hen 'Hills of Hearing and of Sight' (LotRII ch.9), but it was originally intended as one place. Apparently Tolkien toyed with the following elements:

• tir (see TIR- 'watch, guard'), lenited dir
• lhaw 'ears (of one person)' (LAS2- 'listen'), lenited law (cf. S. lhaw < slasū < SLAS- 'ear' (PE17:77))
• las(t) *'hearing', cf. Q. lasta 'listen', lasta 'listening, hearing' (ibid.)
• hen(d) 'eye' (KHEN-D-E- 'eye', N. henn > hên)

He arranged them differently looking for the most suitable solution. All forms may all be easily translated:

• Llawhen - *'ears-eye'
  Most interesting in this example is the spelling ll for unvoiced l according to Welsh orthography; compare the late note on mallorn, which initiated some controversy among scholars:
  [...] in The Lord of the Rings ll is used in the manner of modern Welsh for the medial voiceless l; as in mallorn < malhorn < malþorn < malt 'gold' and orn 'tree' (VT42:27)
  In this context it should be noted that Tolkien explicitely mentions 'medial voiceless l', surely being aware having used the spelling Lhaw instead of Llaw. But the appearance of Llaw in the early drafts may be a hint at the fact that the note from VT42 is not a reinterpretation - ll for voiceless lh was perhaps planned at the outset, although it is not mentioned in the appendix. [5]
• Amon 'hill' (AM2-) and Tirlaw *'sight-hearing'
• Lhawdir *'hearing-sight'
• Lasthen *'hearing-eye' - interestingly not *Lhasthen; a further example of changing Noldorin phonology
• Henlas, Hendlas 'eye-hearing' - the first from drops medial -d- in the -ndl- cluster; both simplify final -st to -s

Finally, Tirmindon and Larmindon are apparently Quenya versions of Amon Hen and Amon Lhaw with tir- + mindon 'tower' (Silm.index) and las- + mindon (rhotacism s > r).

● Kelufain >> Forfain >> Calenbel, Calen-Bel >> Calembel, Cálembel (TI:371,382,384,388)
These are names for a green lawn below Amon Hen, later Calledin and Calembrith (see below), finally Parth Galen (LotRII, ch.10).
The first form seems to mean 'white source' < N. celw 'spring, source' (KEL-) and N. fein 'white' (SPAN-). The final combination ei in the final syllable often appears later (while Noldorin underwent a transformation into Sindarin) as ai, but some variation can be already found in The Etymologies (e.g. N. lhein, lhain 'free(d)' (LEK-)).
One has to wonder, however, how such a name should be applied to a lawn; but the gloss of KEL- is 'go, run (especially of water), flow away downhill' (VT45:19) and might be taken here in a more abstract meaning, as a sloping surface (a green lawn ran down to the water from the feet of Amon Hen (LotRII, ch.10)). But there was also a source: A little spring fell tumbling down and fed the grass (ibid.).
The second form is more difficult. It still contains fain 'white', while the initial element looks like for- 'north' (PHOR-, PE17:18, compare also Forlorn 'North Haven' above), but it makes little sense in this context. Perhaps we are dealing here with N. faur 'beach, shore' (SPAR2- 'strew, spread', VT46:15) and au > o in the compound.
The following forms then appear to mean 'green lawn' literally, containing calen 'bright-coloured = green' (KAL-), later assimilated to the following labial as calem-; and presumably lenited pêl 'tūn, fenced field' (PEL-, VT46:8). Admittedly, a 'fenced field' is not exactly a 'lawn' of a natural surrounding, but I cannot see a better solution here.
In The Lord of the Rings, Calembel is a town by the fords of the Ciril where Aragorn, Gimli and Legolas pass with the army of the Dead (LotRV ch.3, RC:537). It is translated as 'Greenham' with the suffix -ham 'village, homestead' common in English place names, which suits N. gobel, Q. peler, opele 'walled house or village, 'town'' (PEL(ES-)).

● Methen Amon  >> Methendol >> Methedras (TI:391,404)
This is the name of the last great peak of the Misty Mountains; Methedras also in LotRIII, ch.2.
The first two forms include N. methen 'end, final' (MET-, VT45:34) and amon 'hill' (AM2-) or dôl 'head' (NDOL-) or dol 'head, hill' (RC:433), 'head; often applied to hills and mountains' (Silm.index). Interestingly, the adjective is in preceding position in Methen Amon.
Methedras may include the noun methed 'end', as in Methed-en-Glad 'End of the Wood' (UT:452), but an assimilation of methen, -nr- > -ðr- also seems possible (compare Caradras above).

● Orod Thon >> Orod Thuin (TI:420)
This place name appears later in Treebeard's song as Orod-na-Thôn (LotRIII, ch.4).
Both names seem to mean *'pine-mountain' or *'mountain of pines' with N. orod 'mountain' (ÓROT-). The Etymologies give N. thaun 'pine-tree', pl. thuin (THŌN-) as an Ilkorin loan; later Noldorized S. thaun, thôn < thānĭ- or thôn 'pine' < stŏna (RC:384, PE17:81-82). Here, Tolkien obviously decided that the word should have a short vowel o, but kept the plural thuin - compare Duil Rewinion with dol or dôl > pl. duil above.
See also Dorthonion, Dor i Thuin or Dor i thoen 'land of Pines' (PE17:81, RC:384).

● Nan Gurunír (TI:420) < nan, nand- 'valley' (Let:230) + lenited Curunír 'wizard' or 'Saruman' ('man of craft'), most probably from KUR- 'craft'
This appears as Nan Curunír in The Lord of the Rings, but the second word is here lenited in genitival position according to the rules of Noldorin.

● Osgiliath >> Elostirion, Ostirion '= fort' (TI:423)
This changed name of Osgiliath 'fortress of the stars' (Let:347) looks like a translation into Quenya with el- 'star' (EL-) + ostirion < osto 'city, town with wall around' (OS-) + tirion 'watch-tower, tower' (TIR-), translated as 'fort'.
Note that in The Silmarillion, Elostirion is a tower upon Emyn Beraid (western Arnor), where one of the palantíri is kept in the Second Age.

● echain, Ostechain 'Town build again', 'Newbold' (TI:424)
This name was considered as a revision of Osgiliath.
While N. ost 'city, town with wall around' (OS-) is clear to spot here, the word echain 'built again' is somewhat difficult to explain. In Quenya we know several derivatives of the stem EN- 'once more, again'; in The Lord of the Rings it appears in Envinyatar 'Renewer' and enquantuva 'shall refill', but in Noldorin/Sindarin such an element is not attested. Instead, EN in said to be replaced by ATA- in Sindarin because of a clash with ET- 'out' (PE17:167).
Hence, one should rather assume the umlauted prefix ad- 'again' (ATA-, VT45:6) here. It would then cause stop mutation of the following element. KAT- 'shape' with EN. echedi 'fashion, shape', pa. t. echant, echannen (< et-kat-, VT45:19) seems suitable, but the exact derivation is problematic.
Stems like KHAN- 'understand, comprehend', KAN- 'dare', *KHEG-, *KEG- would lead to echain just by suffixing an adjectival ending -jā or -nā. But since final -nt does not change > -n(n), KAT- would require both endings and a syncopation: *et-kat-na-jā > *ekkhadn(a)ja > *echain(n) or maybe an ending *-njā: *et-kat-njā > echain, both very uncertain reconstructions. However, compare the discussion of ruin in Dant-ruin above.

● Bern Lamrach; Tarn Felin; Trandóran (TI:424)
These are earlier names of the bleak hills of Sarn Gebir (also Sarn-gebir, Sern Gebir (TI:283)). Sarn Gebir was originally applied to the highlands later called Emyn Muil 'Drear Hills' (UT:434).
Associated with the early Elvish forms are English names 'Graydon Hills', Grailaws', 'Hazowland' (= ashen-grey land), but the exact interpretation remains a riddle nevertheless.
Bern in the first form could be from BARÁN- yielding words for 'swart, brown', the umlaut being caused by a final -i in the primitive *barani (> Q. varne), compare the experimental N. cern 'ruddy', Q. karne (karani) (VT45:19, KARÁN-). The region further to the North is actually called 'Brownlands' (TI:317,351). Lamrach looks as if it contains N. lham(b) 'tongue' (LAB-) and #rach 'wain' (UT:465) or #rhach 'curse' (MR:373), but all taken together it does not make much sense. I rather suspect that this name is simply uninterpretable, for instance as Eilenach (VT42:19) or Erech (Let:297).
In the second form felin may be related to FURU-, HURU- wiith Early Qenya furin, hurin 'hidden, concealed', FOƷO 'hide, hoard, store up, lay up in secret' with Q. foina 'hidden, secret', fôle 'stealth, a secret, secrecy', fôlima 'secretive' (QL:38-39). Note that muil may have been the adjective 'hidden' from MUY- (whence Q. muina 'hidden, secret', Dor. muilin 'secret, veiled'), before Tolkien re-interpreted it as 'drear', although this is pure speculation (there is deleted N. †muin 'secret' (VT45:36), however) and I cannot tell why this land should be specifically called 'hidden'.
Then we find Tarn in the second name and #Tran- in the third, which suggests that we are dealing with two different syncopations of a stem *TARAN-. In the King's Letter we can later find i Drann as a translation of 'Shire' (SD:129).  #Trann could also be a general word *'province, region' occurring in these early names. Hence, Tarn Felin might be vaguely interpreted as the *'hidden province'.
Tarn also occurs in Tarn Aeluin, the lake on Dorthonion where Barahir and his companions made their lair, and where they were slain (Silm.index), earlier attested in The Lay of Beleriand (LB:339,345,350). But here, it must be English 'tarn' (= a small steep-banked mountain lake or pool) rather than an Elvish word, compare the lines:
[...] above the darkling pines arose
of steep Dorthonion to the snows
and barren mountain-winds, there lay
a tarn of water, blue by day, [...] (LB:335-336)
On the other hand tarn is also attested directly as 'gate' in Early Noldorin and Goldogrin (PE13:153, GL:69, LT1:300), but this does not seem to be appropriate for a wide area.
Also, in the Gnomish Lexicon one can find tram 'cross', trôn '1) a cross, 2) a crossing', trantha- '1) to cross tr. & intr. 2) to mark with a cross', tranc 'criss-cross, crossed' (GL:71). The Qenya Lexicon gives the stem TARA- (QL:89) with similar derivatives - e.g. Q. tarna 'crossing, passage, ford'. If this remained valid, Tarn Felin might be the *'hidden crossing'. An clue which perhaps supports this last interpretation is the town name Tarnost (see below).
In the third form one can identify the suffix -an also seen in Rohan, which is derived from Elvish *rokkō 'swift horse for riding' (Q. rokko, S. roch) + a suffix frequent in names of lands (Let:297). What remains in between is #dór and may be N. dôr 'land, dwelling-place'  (NDOR-, VT45:38), thus Trandóran *'cross-land'? Needless to say that all of this is highly speculative.
Sarn Gebir is later applied to the rapids of the Anduin and translated as 'Stone-spikes' (UT:463) or 'stone-spiked' (RC:327), so called because of the upright stake-like spikes of rock at their beginning, from S. sarn '(small) stone' + ceber 'stake', pl. cebir.
But originally we find notes with kebir, kapir and translations of uncertain reading ?'Linestones', ?'Limestones' or ?'Lonestones' (KEPER-, VT45:20); KOPAR- 'knob, head, tip', N. cobar ?'grave barrow', ?'graibarrow', ?'dunharrow' (VT45:23). This was clearly put down at a time when Sarn Gebir or Sern Gebir was applied to the region of hills, but I cannot identify neither 'grey' nor 'lone' or something similar in one of the earlier names.

Names from The War of the Ring

After the previously mentioned long break Tolkien continued to write on starting with 'The Taming of Sméagol' in April 1944 (WR:77-78, Let:59). The first names below still belong to 1942 (cf. WR:59, note 10), from Sern Erain onwards they are from 1944. The story is now almost brought to its end, the thread with Frodo and Sam is brought until the point when Frodo is captured by the orcs (cf. WR:219-220). The western thread reaches the journey to Morannon after the victory of the Pelennor Fields.

● Rochirwaith, Rochirchoth, Rochirhoth, Rochann, Rohann, Rohirrim (WR:22)
● Rochiroth (WR:168) - see Rohiroth, Rochiroth (RS:440)
● Rohar? (WR:137)
In Appenidx E to The Lord of the Rings we learn that ch [...] was weakened to h in the speech of Gondor, and that change has been recognized in a few names, such as Rohan, Rohirrim and that Rohan is a modern form of Rochann < Rochand (cf.also Let:297).
It seems that the development of this conception is mirrored in the note with experimental forms in the first line above. Rochirwaith is clearly from Rochir 'horse-lord' (Let:144) and gwaith, literally 'manhood, also used = man-power, troop of able-bodied men, host, regiment' (gweith under WEG-, VT46:21), frequent in names of peoples as Forodweith 'Northmen'. Rochirchoth shows suffixed hoth 'host, crowd, frequent in people-names as Glamhoth (KHOTH-), lenited to -choth; while lenition is absent in Rochirhoth.- compare Glamhoth 'the barbaric host' (GLAM-), i ngaurhoth *'the wolf-host' (LotRII, ch.4; ÑGAW-).
Rohar appears with a question mark in connection with other notes. It may signify *'hose-king' from roch 'horse' (Let:144) and âr 'lord or king of a specified region' (TĀ-,TAƷ-, cf. ƷAR-) or rather *'horse-lord' equivalent to Rochir, compare Gwaewar 'the Windlord' rather than *'Wind-king' above.

● Fuilas >> Marlas >> Romloth >> Galenas (WR:38)
These are earlier names of galenas, the name of pipeweed in Noldorin/Sindarin (see also LotRV, ch.8).
Fuilas clearly contains N. lhass, S. las(s)  'leaf' (Let:211, LAS1-). The first element must be from PHUY- with N. fuin 'night', Q. Fui, Hui 'Night', fuine, huine 'deep shadow'. Indeed, tobacco belongs to the Solanaceae or nightshade family and this seems to be translated here by *'night-leaf'.
Marlas is more difficult - while suffixed -las 'leaf' can be also observed in this word, the only suitable element mar- appears in a first translation of the name Hamfast (= 'home-fast, home abiding') as Marthanc (SD:117) with mutated tanc 'firm' (TAK-) and thus obviously mar- 'home' < MBAR-, despite the usual change mb- > b- as in N. bár 'home' (VT45:33). Perhaps mar- is from a parallel root *MAR- or underwent some kind of influence. However, it is difficult to see how Marlas *'home-leaf' relates to pipeweed. Perhaps it is simply a misreading for *Morlas with mor-'night, dark' from MOR-. This would be a variation of Fuilas.
Romloth contains N. lhoth 'flower' (LOT(H)-) or S. loth 'flower' (Silm.index), 'inflorescence, a head of small flowers' (VT42:18) and as the initial element probably N. rhom 'horn, trumpet' (ROM-) with voiced r- according to the later phonology of Sindarin. The shape of tobacco blossoms may indeed remind of trumpets.
There are two different explanations of galenas from Tolkien in PE17:100. According to the first the translation '= nicotiana, Gondor Sindarin for 'pipeweed'' is given, followed by the form galen and Quenya forms alanasse, ale that are added without translation. This seems to point towards GAL2- 'grow, spread, increase' (or 'grow/flourish' in PE17:153) with N. galas 'growth', N. galo 'to grow'. Perhaps the formation goes via adjectival *galanā > *alana 'grown' and the noun suffix -sse in Quenya, while Sindarin substitutes galen < *galinā medially (by analogy). However, the mentioning of 'pipeweed' inevitably leads to the suspicion that 'pipe' is contained somewhere within the words - this can only be #nasse or #asse (N./S. #-as), but it cannot be conceivably related to any known root.
The second explanation cites the element nes- 'sweet smelling' > *S. ne 'scent'. Gondor Sindarin galenas is said to be from galanes (hence probably a corruption by the Gondorians) and would thus originally mean *'sweet smelling plant', cognate of Q. alenesse, alanesse < ale, #ala 'plant' + #nesse *'sweet smelling'. Compare the description in the Prologue to The Lord of the Rings:
The Men of Gondor call it sweet galenas, and esteem it only for the fragrance of its flowers.

● Galbedirs (WR.47)
● Lamorni, Ornómar (WR:50)
● Ornómi (WR:52)
● Huorns (WR:30)
These are earlier names for the trees of Fangorn which are able to move. In English they are referred to as 'Talking Trees' - according to Treebeard they still have voices, and can speak with the Ents [...], but they have become queer and wild (LotRIII, ch.9). In the final version they are called Huorns (ibid., WR:30, RC:425).
An analysis of the earlier forms was done by Philipp Marquart in Lambengolmor message #913 and I shall give a summary:
Initial gal- of Galbedir must be derived from GÁLAD- 'tree' or GALA- 'thrive (prosper, be in health - be glad)'. The suffix -bedir is most probably lenited #pedir *'speaker' from KWET- 'say'. This word itself could contain †dîr *'man' (DER-) surviving chiefly in proper names [...] and as agental ending. The example ceredir 'doer, maker' is associated with this note. But on the other hand the name Gelir as a translation of 'Merry' (SD:129) seems to have been formed from GYEL- (yielding e.g. N. gell 'joy, triumph') with the sole suffix -ir. Thus there appear to be two possibilities: *ped-dîr > #pedir or *kwet-iro > #pedir *'speaker'.
Lamorni is Quenya; it seems to contain #lam- 'sound' (LAM-) and orni, plural of orne 'tree, high isolated tree' (ÓR-NI-, UT:168).
Ornómar consists out of orne (#orn-) and óma 'voice' (OM-), pluralized with -r.
The same orne or #orn- is part of Ornómi. The latter element could be most easily explained here as a direct plural #ómi from the stem OM- as bal-ī > Vali from BAL- beside Valar; but it could also be that Tolkien imagined a different formation - *óme 'voice', pl. *ómi instead of óma, pl. #ómar. Compare for instance Q. nóte 'number' (NOT-).
For Huorn PE17:86 cites a couple of scribbled forms. At first Q. hú, S. hû noun ?'hound' (?'heart') < khōgo - this seems to be ultimately related to KHUG- 'bark, bay' with Q. huo 'dog', N. hû. The application of this root here may be due to the fact that the Huorns do not speak a proper language (lambe), but rather an animal-like speech. Also associated are khōn-, *Q. hôn, *S. hûn *'heart' (compare KHŌ-N- 'heart (physical)' with the same derivatives) and hō ?'spirit, shadow' or ?'speak, show'. This application might express something like 'trees that are alive' - who have a spirit or heart.

● Carandrian (WR:55) < N. caran 'red' (KARÁN-) + #rian, perhaps a suffix similar in role to #-ian as in perian 'halfling' (Let:347, PER-) and epenthesis of -d- (cf. Celebrindrath above)?
This is the name of a rowan-ent (thus the name refers to the red berries), friend of Bregalad.

● Angost (WR:72) < N. ang 'iron' (ANGĀ-) and ost 'city, town with wall around' (OS-).
This was a passing name of Isengard, earlier Angrobel, later Angrenost (see above).

● Mount Tor-dilluin >> Mount Mindolluin (WR:77, 80)
This is the mountain at whose foot Minas Tirith was built.
Tor-dilluin probably contains N. taur 'vast, mighty, overwhelming, awful - also high, sublime' (TUR-), shortened to tor- in the compound. The second part can be interpreted as lenited tilluin, iteslf consisting out of N. till, tild 'horn' (TIL-) or S. til 'point, horn' (Silm.index), S. till 'tine, spike, point' (PE17:36) - frequent in mountain-names - and S. luin 'blue' (Silm.index). Thus altogether it may mean *'Mighty Bluehorn'.
Mindolluin is translated as 'Towering blue-head' (RC:439) in the unfinished index and might contain mindon 'tower' (MINI-) at the first glance, but The Silmarillion index states that it contains dol 'head; often applied to hills and mountains' and #min- which occurs in other words referring to isolated, prominent, things, thus #mindol *'isolated, prominent mountain'.

● Mithond (WR:77,80)
● Mithrond (WR:80)
These names appear for later Mithlond, the 'Grey Havens' and Mithrond was changed to Mithlond.
Mithond may be just a slip, as suggested by Christopher Tolkien. If not, it appears to mean *'grey stone' with N. mith 'grey' (MITH-) and lenited N. gonn 'a great stone or rock' (GOND-). The application of this name to a haven is unclear, but it might refer to coastal cliffs.
Mithrond appears to contain N. rhond, rhonn 'cave' (ROD-) as the second element. If this is not a slip either, it might be referring to caves in coastal mountains. But on the other hand, Ilk. rond means 'domed roof' and later we find S. rond 'vaulted or arched roof, or a large hall or chamber so roofed' (Silm.index). Thus, r(h)ond might somehow (pars pro toto?) denote here that the harbour shields the ships from outer storms.

● Sern Erain >> Sarn Aran (WR:98) 'the King Stones = the Gates of Sarn Gebir'
● {Sarnel Ubed. Ennyn. Aran} Taur Toralt {Sarn Torath} Annon Torath. Aranath (WR:132)
● Sairn Ubed (WR:132)
These are earlier names of Argonath < arn(a)gon-ath 'the group of (two) noble stones' (Let:347), earlier Sern Aranath (see above). The sequence in the second line occurs in a brief outline for the chapter 'Of Herbs and Stewed Rabbit', which was put down in a hardly legible handwriting. Frodo and Sam see a stone figure on their journey through Ithilien which reminds them of Sern Aranath. In the The Lord of the Rings they encounter it in the chapter 'Journey to the Cross-roads':
The brief glow fell upon a huge sitting figure, still and solemn as the great stone kings of Argonath. The years had gnawed it, and violent hands had maimed it. Its head was gone [...]
One can only guess that the singular forms among these hastily written names refer to the broken statue in Ithilien and the other ones to the two statues at the Anduin.
Sern Erain is formed with sern pl. of sarn 'stone as a material; or as adj.' (SAR-) and erain, pl. of aran 'king' (ƷAR-, Let:347). The following Sarn Aran shows the same words in the singular.
Sern Ubed was added later to the same page with 'denial' in brackets (WR:137). This must be a translation of ubed, which can be hence analyzed as containing the negative suffix u- (UGU-, UMU-) and the lenited element -ped from KWET- 'say'. For a derivation of this kind compare N. úan 'monster' < ūbanō (BAN-), the vowel becomes shortened in uanui 'monstrous, hideous' or úgerth *'tresspasses' < carth 'deed' (KAR-). In The Etymologies we can find pencilled N. ú- 'un, bad-' (VT46:20). But all these forms imply a reversal, which is not the derivation of 'deny' from 'say'. However, one can also find the struck through N. mû 'no' from MŪ- 'not, no' (VT45:35) which looks related to UMU-. This would explain ubed *'saying no = denial'. The name *'Stones of Denial' most probably refers to the fact that the two statues hold their hands raised in a gesture of refusal.
The word sarnel instead of sern or sairn is formed with the suffix -el, which also occurs in N. gandel, gannel 'a harp' (ÑGAN-, ÑGÁNAD-), findel, finnel '(braided) hair' < sphindele (SPIN-) or nelladel 'ringing of bells' (NYEL-). Here its role might be the formation of *'stone-figure, statue'. See also Dant-ruinel, Nen-uinel above.
From the following notes Ennyn Aran should probably belong together, meaning *'Gates of the King' with pl. ennyn of N. annon 'great gate' (AD-). As the statues mark the northern border of Gondor on the Anduin, they are also in a way a gate to its land.
Taur in the next form Taur Toralt may either be N. taur < tāro 'king: only used of the legitimate kings of whole tribes' or the adjective taur < taurā 'vast, mighty, overwhelming, awful - also high, sublime' (TUR-) and Toralt most probably contains taur intially with au > o in the compound. The second element -alt could be the cogante of Q. alta 'large, great in size' (ÁLAT-). As a whole toralt may mean *'king-large = large king' (referring to the size of the statue) or be the compound of two adjectives *'large and mighty'.
Sarn Torath probably means *'Stone of the Kings' with taur 'king' in the class plural, similar to Sern Aranath.
Annon Torath must then be the *'Gate of the Kings' with annon 'gate' in the singular.
And finally, Sairn Ubed shows a different plural formation of sarn to sairn instead of sern.

● Neleg Thilim >> Neleglos 'the White Tooth' (WR:106)
This is an older name of Minas Ithil (internally). An analysis is given in Lambengolmor message #921 by Philipp Marquart.
Both forms transparently contain N. neleg 'tooth' (NÉL-EK-). The adjective thilim looks like being derived from THIL-, a root parallel to SIL- 'shine silver'. So one can isolate the adjectival ending -im, which already occurred in #mithim *'hoar, grey', Nenvithim 'Hoardales' (TI:114) - see above.
Neleglos likely contains N. gloss 'snow, also adj. snow-white' (GOLÓS-); later we find S. gloss '(dazzling) white' from (G)LOS- (VT42:18).

● Mornennyn >> Morannon (WR:112)
● Ennyn Dûr >> Morennyn >> Mornennyn (WR:113)
● Kirith Naglath 'Cleft of the Teeth' (WR:137)
These are preceding forms of the Black Gate of Mordor, in The Lord of the Rings it is called Morannon (LotRIV, ch.3).
The elements involved here in several combinations are N. morn 'black', cognate to Q. morna or N. †môr 'black', cogante to Q. more < mori (MOR-, VT45:35; Let:347) (or else just a prefix mor- directly from the stem); and N. annon, pl. ennyn 'great gate' (AD-). Many European languages put 'gate' or 'door' into the plural (a gate is naturally formed out of two halves), e.g. Slavic vrata, vorota is always used in this plural form. It seems that Tolkien for some time imagined the same to happen in Noldorin/Sindarin.
The form Ennyn Dûr is peculiar, since the latter part of it is the adjective dûr 'dark, sombre' (DOƷ-, DÔ-) which does not agree in number with the pluralized ennyn. The usage of ennyn as singular despite being grammatical plural seems unlikely, since Nelig Morn and Naglath Morn occur at the same time (see directly below). It may be just a slip (Nelig Morn is also immediately changed, Naglath Morn deleted) or perhaps dûr < *doʒrē is here meant to be the noun *'blackness, darkness' and is put into genitival position.
Kirith Naglath is from cirith 'cleft, ravine, defile' < kir- 'cut, cleave' (RC:334-335, Silm.index, cf. KIR-) and naglath collective pl. 'teeth', for which see directly below.

● Nelig Morn Mel >> Nelig Myrn 'Teeth of Mordor' (WR:113,122)
● Naglath Morn (WR:122)
These are epithets for the two towers guarding the entrace to Mordor, earlier Gorgos and Nargos (see above), later Narchost and Carchost 'Towers of the Teeth' in The Lord of the Rings (LotRVI, ch.1).
Nelig is clearly the plural of N. neleg 'tooth' (NÉL-EK-). Naglath is obviously formed with the class-plural suffix -ath, so that the basic form can be guessed *nagl, *nagol < *nakla from NAK- 'bite'; compare N. magl, magol 'sword' from MAK- 'sword' or 'fight (with sword)'. Mel is probably the beginning of an unfinished word.
The following adjecitve is N./S. morn 'black' (MOR-, VT45:35; Let:347), interestingly in the singular, although qualifying pl. nelig, naglath. Elsewhere Sindarin and Noldorin adjectives are seen to agree in plural, so that this might be just a slip. In any case Nelig Morn has been immediately changed to Nelig Myrn with pl. myrn, while Naglath Morn was subsequently struck out and not replaced.

● Hebel Dúath >> Ephel Dúath (WR:122)
These are Elvish names for the 'Mountains of Shadow', a mountain range marking the southwest of Mordor, translated as 'Fence of shadow' in the unfinished index (RC:457) with ephel 'outer fence' < et-pel < pel- 'go round, encircle', compare Et-pele > Eppele > Ephel (WR:137), ephel < eppel < et-pel 'outer wall or fence' (PE17:65), PEL- 'revolve on fixed point' in The Etymologies; and Dú(w)ath 'night-shade' (DOƷ-, DÔ-).
If the earlier hebel is derived from PEL- as well, it needs a prefix *he- instead of et-, but nothing like this is attested. Instead we may assume that the underlying root is √khep- 'retain, keep, do not give away or release, keep hold of' (VT41:6, PE17:157) - as also in the Sindarin verb ú-chebin lit. *'I do not keep' (LotR App.A) - with the meaning 'fence' derived from it. For the formation compare keglē > S. cail 'a fence or palisade of spikes and sharp stakes', thus perhaps khepelē > hebel with ómataima.

● Andabund >> Andrabon >> Múmund (WR:136)
● Múmar, Mâmuk (WR:136, Let:66)
These are various experimental names for the oliphaunt of Harad, ultimately mûmak was chosen (LotRIV, ch.4).
The Etymologies give MBUD- 'project', N. andabon, annabon 'elephant', lit. 'long-snouted' with the first element anda-, anna- meaning 'long' (ÁNAD-, ANDA-) and the adjective #bon(d) < #mbundā 'snouted'.
Andabund seems to be formed with the noun bund, bunn 'snout, nose, cape' instead and leaves medial -nd- unchanged (compare Mindolluin, Gondolin, Glorfindel (Silm.index)). Andrabon is formed with #bon again, but shows a different adjective initially, probably from primitive *and'rā with adjectival -rā.
Múmund looks like a Noldorin/Sindarin word as well, with bund being lenited to *mund (cf. N. bâr 'house' > i·mâr (PE13:120), initial mb- archaically). But then the element mú- is difficult to explain. In any case the later mûmak is from the speech of the Easterlings (PM:79) and so are most probably its predecessors múmar and mâmuk. Múmund may have been conceived as an alien word as well with just a coincidental resemblance or perhaps it might be a mixed compound.
Compare also the pseudo-Elvish translations Aramund 'Kingly bull', Tarmund 'Noble bull', Rasmund 'Homed bull', Turcomund 'Chief of bulls' (Let:345) that can neither be Quenya nor Sindarin phonologically.

● Hebel ?Orolos >> Hebel Uilos Nimr[?ais] (WR:137)
● Ephel Nimras >> Ephel Nimrais, Ered Nimrath (WR:137,156)
● Ered Nimras (WR:167-168)
At this point the 'Black Mountains' (see Ered Myrn, Eredvyrn, Mornvenniath above) were changed to 'White Mountains' with several Elvish experimental forms. The ultimate form was Ered Nimrais 'White-horn mountains' (UT:436).
Hebel did already occur in Hebel Dúath as a name for a mountain range (also Ephel, see above). The form ?Orolos shows oro- from ORO- 'up, rise, high etc.', here either describing something like *'height, mountain peak' or being plainly 'mountain' as in Orofarne 'mountain-dwelling' (Let:168). The second element is presumably lenited N. gloss 'snow, also adj. snow-white'. Thus Hebel Orolos *'Range of snow-white peaks'.
Uilos is a name elsewhere used for the Valinorean mountain Oiolosse 'Ever-snow-white' (OY- UT:55, Rgeo:74), the initial element ui- comes from OY- 'ever, eternal'.
Nimr[?ais], if read correctly, is the plural of nimras 'white-horn' < N. nimp, nim 'pale' (NIK-W-) or nim 'white (usual word)' < Telerin nimbĭ, alteration of nimpĭ < *ninkwĭ < NIK- (PE17:49,168) + *ras(s) or lenited ?rhass 'horn especially on living animal, but also applied to mountains' (RAS-, VT46:10), of uncertain reading in The Etymologies (S. rass 'horn' in PE17:36).
Interesting is Nimrath, which seems to contain rath 'course, riverbed' (RAT-). On the other hand a later source cites rath- 'climb' (UT:255) and rath, applied in Númenórean Sindarin to longer road-ways and streets of Minas Tirith (which were inclined), as Rath Dínen 'the Silent Street'. But a root meaning 'climb' may also yield *'top, mountain peak'. Notable in this context is earlier Caradras 'Redway' (see above) with the roles of r(h)as and rath reversed. So it may be that Tolkien just wavered concerning the application of these two words.

● Máraher 'the good steward' (WR:153) < Q. mára 'useful, fit, good (of things)' in The Etymologies (MAG-), later 'good, as it should be, in right or proper form or state, in health, well' < √MAGA (PE17:162, cf. VT42:33-34) + -her 'master' < KHER- 'rule, govern, possess'
This is a passing name of later Mardil, the second stewart of Gondor.

● Shorab, Shorob (WR:153)
● Forlong >> Fornold >> Incânus (ibid.)
● Olórion >> Olórin (ibid.)
These are earlier versions of Gandalf's names in the different regions. In The Lord of the Rings he says:
Many are my names in many countries, he said. Mithrandir among the Elves, Tharkûn to the Dwarves; Olórin I was in my youth in the West that is forgotten, in the South Incánus, in the North Gandalf; to the East I go not.' (LotRIV, ch.5)
Mithrandir and Sharkûn at once appear in this earlier draft. Especially interesting are Gandalf's names Shorab, Shorob in the east, whereas he does not go there in the final version. This might be how he was called by the orcs.
Christopher Tolkien writes: On Gandalf's names 'in the South', Forlong changed to Fornold I can cast no light. He then states that Forlong became the name of the lord of Lossarnach (LotRV, ch.1) and in the Appendix F Forlong is mentioned among names of of forgotten origin, and descended doubtless from days before the ships of the Númenoreans sailed the Sea. Thus it appears that Tolkien just had a suitable name-shape in mind, without any specific meaning attached to it and changed its application. Fornold is most probably of the same sort. Note however Tom Bombadils Scandinavian name among the Dwarves Forn '(belonging to) ancient (days)' (RC:128). On the connection of Incánus with Latin see UT:400; an Elvish etymology includes Q. incānus(se) 'mind mastership' (PE17:88), in-kāno/u 'mind master' (PE17:155) < IN-ID 'mind, inner senses', KAN- 'lead'.
For the interpretation of Q. Olórin see UT:396-397, the earlier Olórion just has a different name-formative suffix -ion.

● Elenarda, Kalen(arda) (WR:155,167)
● Kalinarda (WR:156)
● {Kalin} Calenardan >> Calenardhon (WR:168)
These are earlier ancient names of the region which was in the Third Age renamed to Rohan, ultimately Calenardhon 'The Green Province' (Silm.index).
Elenarda is Quenya 'star-realm, upper air or sky' (VT45:16) < Q. elen 'star' (EL-) + arda 'realm' (ƷAR-). The new reference of this name to the region of Rohan is difficult to explain. Perhaps it is because of the stars becoming better visible when standing amidst such a plain? Or might this region be seen to form a star-like shape on the map?
Kalen(arda) was written above Elenarda and is a peculiar form - N. calen 'bright-coloured = green' (KAL-) must come from *kalinā (see Q. kalina 'light (adj.)') with a-affection i > e. But a-affection is not observed in Quenya, while arda is clearly a Quenya name and cannot be Noldorin/Sindarin. Is it just a slip of Tolkien's or an error of the Númenoreans? In any case the following version of this passage (Faramir's description of the history of the Dúnedain) has Kalinarda, the Quenya form one would expect.
Calenardan is obviously from calen + #ardan *'realm'. According to The Etymologies the stems ƷAR- 'have, hold' and GAR-, GARAT-, GARAD- were much blended in Eldarin. A Noldorin derivative is ardh 'realm', while ardan seems to show the suffix -an denoting a region as in Rochann > Rohan 'Hippia' (Let:144). Thus ardan may here symbolize *arðan - see UT:267 on this transcription matter - unless ƷAR- blended here with GARAT- rather then GARAD-.
Calenardhon contains ardhon *'realm' now with a different suffix -on (cf. Arðon in PM:348).

● Nargalad, Anngalad, Carangalad, Henneth Carandûn, Henneth Malthen, Henlo Naur, Henlo n'Annun, Hennuil n'Annun (WR:164)
These are earlier names of Henneth Annûn 'Window of the Sunset' (LotRIV, ch.5), a waterfall in Ithilien. It screened the entrance to a cave the rangers were using as a hiding-place. The cave faced west, so that whenever the sun would go down its light would be broken into many flickering beams of ever-changing colour, hence the names:
Nargalad should mean *'red light' with nar- 'red' from NAR1-, whence also N. narw, naru 'red', compare Narosîr 'Redway' and Narodûm 'Red Vale' above, Nardol below. The second element is then most probably lenited N. calad 'light' (KAL-).
Compared to the other forms Anngalad may conceivably signify *'light of the west', but then it would contain annûn 'west' (NDŪ-) shortened to ann- with the original derivation being hardly recognizable. Another theoretical possibility would be N. and, ann 'long' (ÁNAD-, ANDA-) as the initial element, but I have no clue why the waterfall should be called *'long light'.
Carangalad is again *'red light' with N. caran 'red' < KARÁN- and calad 'light'.
Henneth Carandûn is probably *'window of the red west' with henneth *'window' according to the translation of later Henneth Annûn. Henneth is perhaps derived from KHEN-D-E- 'eye' with the suffix -eth (cf. N. meleth 'love' < MEL- 'love (as friend)') denoting a noun, although such nouns usually have an abstract meaning. Carandûn must be from caran 'red' and andûn or dûn 'west'.
Henneth Malthen is the *'golden window' with N. malthen 'of gold', analogical for mallen (SMAL-).
In Henlo Naur the first word seems to contain N. henn or hên 'eye' (KHEN-D-E-) and lô, earlier 'a pool, lake' (GL:54), later attested as 'fenland' (UT:263, VT42:9-10) from LOG 'wet (and soft), soaked, swampy etc.', while naur means 'fire' (NAR1-). Thus we can tentatively suspect a meaning like *'wet eye/sight of fire'.
In Henlo n'Annun the word naur is exchanged by annun 'west', for some reason with a short vowel (probably Tolkien was writing hastily and left eh diacritics out). The prefixed n' is the shortened genitive marker na, also appearing in later mír n'Arðon 'Jewel of the World' (PM:348) and frequently in Noldorin place names like Taur-na-Faras 'Hunters' Wold' (SPAR-, LR:450), Taur-na-Fuin 'Forest of Night' (LR:450), Dor-na-Fauglith 'Land of Gasping Thirst' (PHAU-, LR:423); also in the form an in Rath a Chalardain, Ernil a Pheriannath, see below.
Finally, Hennuil n'Annun contains hennuil, probably formed again with N. henn 'eye'. The element #uil may be from ULU- 'pour, flow', presumably signifying *'pouring water, waterfall'. Although the Noldorin of The Etymologies shows another development: ulyā > N. œil, eil 'it is raining' [4], compare the later S. ruin 'fiery, red' < *runjā < √RUN  'red, glowing' (PM:366), cognate to Q. runya.

● melinon >> lebendron >> lebethras (WR:176)
These are earlier names of the tree lebethron. A detailed analysis of the last forms was made by Didier Willis, see [7].
For lebendron he suggests a compound with l(h)eben 'five' (LEP-) and syncopated doron 'oak' (DÓRON-), discussing several primitive variants; and for lebethras a formation from lepet- > l(h)ebed 'finger' (LEP-, LEPET-, later: VT47:23-24) similar to gonathras 'entanglement' (NAT-), thus perhaps meaning *'something made out of fingers, enfingerment'. The final form lebethron then could be either derived with the name-formative suffix -ron (although one might have expected *lebedron, as hadron 'thrower' (KHAT-)) or be an ancient compound *lepet-doron > *lepetdron- > *lepettron- > lebethron. Thus, as he argues, all these names may refer to the 'finger-oak', so called because of its leaves resembling a hand with five fingers.
PE17:89 gives lebethron 'Gondor hardwood, kind of a tree' < lepeth-ron, associated with Q. lepse which is 'finger' under LEP-, LEPET-. The cognate of lebeth is Q. lepetta probably because of its leaves (like chestnut) [were] shaped like a fingered hand. But for the final part two different interpretations are given. According to the first oron 'tree' becomes -(o)rŏnō in compound. According to the second lebethorn > lebethron by association with √RUN- 'rub, grind, smooth, polish' because lebethron was used as word for the wood which took a high polish.
I may add on the first name melinon. Here, one may recall melin 'friends' from The Return of the Shadow (see above), but a tree with such a name seems doubtful. Instead we may be dealing here with N. malen 'yellow' < ON malina (SMAL-). The medial -i- would not undergo a-affection in the compound and instead cause i-affection of the preceding -a-, as in Melthinorn (ibid.) with N. malthen 'of gold', analogical for mallen < *maldina and N. orn 'tree, high isolated tree' (ÓR-NI-). The suffix -on is often seen to form names in Noldorin and Sindarin, thus *malin-ondō > melinon.

● Terch Ungol 'the Spider's Lair' >> {Te} Torech Ungol 'the Spider's Hole' (WR:202)
These are Noldorin/Sindarin names for Shelob's lair. Torech Ungol is mentioned in LotRIV, ch.9.
For ungol 'spider' compare QL:12, root GUNU with Q. ungwe 'spider', esp. Ungwe 'the Gloom-weaver', Gn. gung 'spider' (also GL:43). In The Etymologies UÑG- was reinterpreted as *'gloom' with Q. ungwe 'gloom', Ilk. ungol 'darkness, ungor 'black, dark, gloomy', in N[oldorin] not used except in name Ungoliant. But LotR App.E already gives Q. ungwë 'spider's web' again. Thus one can model *uñg-la > *ungl > ungol 'spider'.
Terch 'lair' may be related to TER-, TERES- 'pierce', whence e.g. Q. tereva 'fine, acute', teren 'slender'. The primitive form can be guessed as *ter'kā. One could imagine that the meaning 'pierce' evolved into 'pierced passage, lair, tunnel'. Compare also Gn. tereg, terch 'a worm' (GL:70), most probably related to TERE- with similar derivatives Q. tereva 'piercing, acute, shrill, sharp', teren 'lissom, lithe', teste 'small worm' (QL:91). 'Worm' seems to be derived here in a very similar way - a worm is not only slender, but also pierces its way through the ground.
Torech 'secret hole, lair' comes from √TOR- 'secrete, hide' and #ech < ekka 'hole' (PE17:188)

● Zaglûn, Gazmog, Nagur-Danlo, Nûzu, Yagûl (WR:212-213)
● Ballung (WR:224)
● Gazmog, Uftak Zaglûn, Ufthak (WR:225)
● Dushgoi (WR:216, 226)
Dushgoi is the Orcish name for Minas Morghul, the other forms are personal Orcish names.

● Amon Thorn, Elenach, Nardol, Penannon, Orodras, Mindor Uilas (WR:232-233)
These are earlier names of the Beacons in Anórien. Their final names were analyzed by Tolkien himself in the essay 'Rivers and Beacon-hills of Gondor' (VT42:18-21, UT:319).
Amon Thorn corresponds to later Amon Dîn 'the silent hill' and could mean *'eagle hill' with N. amon 'hill' (AM2-) and thorn *'eagle' as unflected genitive - The Etymologies give thôr and unsyncopated thoron (THOR-, THORON-), but compare Arathorn (Let:347). Another alternative would be *'fenced hill' with thorn < *thurnā from THUR-, which yields N. thoren 'fenced' < tháurēnā. A third possibility would be thorn *'stiff' < *stār(a)nā < STAR- or STARAN-, whence Ilk.thrôn 'stiff, hard' < starāna. And finally a fifth one with thorn 'steadfast' < √THOR (PE17:113).
Elenach corresponds to later Eilenach, which is probably an alien name; not Sindarin, Númenórean or Common Speech (VT42:19). The same is most probably true for Elenach.
On Nardol 'fiery head' see UT:455. It is obviously composed with nar- 'red' or 'fire' from NAR1-, whence also N. narw, naru 'red', naur 'fire'; and dôl 'head' (NDOL-) or dol 'head, hill' (RC:433), 'head; often applied to hills and mountains' (Silm.index), S. dol (doll) (PE17:36). Compare also Narosîr 'Redway', Narodûm 'Red Vale', Nargalad *'red light' above.
Penannon looks as if it contains N. annon 'great gate' (AD-) and as the initial element N. pend, penn 'declivity', later penn 'slope' (RC:525). But *'sloped gate' as a name for a hill seems doubtful. Perhaps one should rather suspect N. and, ann 'long' (ÁNAD-, ANDA-) here with the name-formative suffix -on, thus *'Long-sloped one'. Medial -n- rather than -nn- despite penn may be explained by analogy, as a formation from the further simplified *pen. Compare arn(a)gon-ath > Argonath (Let:347) rather than *Argonnath. But see also Erelas (VT42:19) which also seems to contain two interpretable elements er- 'single' and las(s) 'leaf', but is not Sindarin.
Orodras looks interpretable again, with N. orod 'mountain' (ÓROT-) and ?rhass 'horn especially on living animal, but also applied to mountains' (RAS-,VT46:10), of uncertain reading in The Etymologies (S. rass 'horn' in PE17:36), thus simply *'mountain-peak'.
In Mindor the first element is probably related to MINI-, whence Q. minda 'prominent, conspicuous', N. minnas, mindon 'tower'. In the Silmarillion index 'tower' is S. minas and the same stem occurs in other words referring to isolated, prominent, things - in this case a hill. The suffix is probably #-or 'hill, mountain' from ORO-, compare Erebor 'The Lonely Mountain' from The Hobbit and Brandor above. Thus mindor could be something like *'isolated mountain'. Uilas could mean *'ever-leaf' or *'ever-leaved' with ui- as in N. Uilos 'Everlasting snow' (OY-) and l(h)as(s) 'leaf' (LAS1-, VT42:19). Mindor Uilas corresponds to later Halifirien (modernized Old English 'Holy Mountain'), on whose slopes the Firien Wood grew.

● Harns (WR:253) < N. harn 'southern' (KHYAR-) + English plural -s; a term for the Haradrim, compare Barangils above

● Taurost 'High City' (WR:260) < N. taur 'vast, mighty, overwhelming, awful - also high, sublime' (TUR-) + N. ost 'city, town with wall around' (OS-)
In The Lord of the Rings Minas Tirith is called the 'High City', but judging from the context (Gandalf and Pippin enter Minas Tirith, ride up to the 6th court, dismount and pass into the High City), this name seems to refer here just to the highest part of the city containing the cidel.

● Rath a Chalardain 'Street of the Lampwrights' (WR:287)
● Ernil a Pheriannath (ibid.)
The first name is a street of Minas Tirith, later Rath Celerdain (LotRV, ch.1) < rath 'street (in a city)' from rath- 'climb' + celerdain, pl. of calardan 'lampwright' from calar 'a portable lamp' (RC:523, UT:255, PE17:96). Compare RAT-, KAL-, TAN-.
Ernil i Pheriannath 'Prince of the Halflings' (Rgeo:75) < ernil 'prince' (UT:428), probably arn 'royal' (Let:347) + suffix -il; and periannath, coll. pl. 'the Hobbits (as a race)' from perian 'halfling' (cf. PER-) with nasal mutation after the plural article.
In both names the genitive particle an < √ANA/NĀ (PE17:146) is used. It causes nasal mutation to the following words: celerdain > a chelerdain and periannath > a pheriannath. Another attestation of it is in an objective role: dagnir an Glaurung 'Slayer of Glaurung = he who slew Glaurung' (PE17:97), but obviously no such specification is made in these early forms.
A discussion of the derivatives of √ANA/NĀ mentions that na, before vowels nan with nasal mutation, means "with" in sense of possessing, provided with, especially of characteristic feature (PE17:147), so i·arben na megil and 'The Knight of the Long Sword'. This could apply to Rath a Chalardain - the lampwrights are a characteristic feature of the street. See also Henlo n'Annun, Hennuil n'Annun above.
Also notable is calar in the earlier plural calardain compared to later i-affected celerdain.

● Faragon >> Orondil (WR:281,288)
These are passing names for the father of the stewart Mardil, later Vorondil (LotR App.A).
Faragon is apparently formed with the same prefix fara- as in Faramir. It is probably from SPAR- 'hunt, pursue' (whence N. faras 'hunting, feredir 'hunter'), since Faragon father of Mardil hunted the wild oxen of Araw [= Oromë] in the far fields of Rhûn. The horn which Boromir bore descends from him. The suffix -gon < N. caun 'valour' (KAN-) appears in many other names.
Orondil is Quenya and seems to mean *'lover of mountains' with Q. oron 'mountain' (ÓROT-) and -(n)dil which implies 'devotion', 'disinterested love'  (Silm.index) and is usually translated as 'lover' (UT:401, NIL-, NDIL-).

● Beren, son of Turgon >> Barathil >> Barithil (WR:282,287) >> Berithil son of Baranor (WR:288)
● Bergil (WR:287)
● Duilas, Garathon (WR:283) >> Targon (WR:288)
● Ramloth son of Thalion >> Gwinhir son of Thalion (WR:284-285, 293)
● Asgil-Golamir (WR:287)
● Inram the tall (WR:252)
● Benrodir (WR:252)
● Nosdiligand, Northiligand (WR:252,266)
● Dirgon (WR:316-317)
These are names of men from Gondor. In The Lord of the Rings Berithil is 'Beregond son of Baranor', his son is named Bergil from the beginning. Targon is his friend in the buttery. Ramloth or Gwinhir is a boy Pippin meets while walking through the city. Asgil-Golamir is later named Golasgil and is the lord of Anfalas coming to Minas Tirith with his soldiers. Dirgon is later named Hirgon (LotRV, ch.3), he is a messenger from Gondor to king Théoden.

• For N. beren 'bold' as a proper name see BER-.

• For Turgon see TUR- and KAN-, at that time it seems to be a compound of tūr 'victory' and caun 'valour'.

• Barathil may contain N. bara 'fiery, also eager' frequent in masculine names as Baragund, Barahir (BARAS-). The second element #thil is most probably from THIL-, variant of SIL- 'shine silver'. Alternatively -il could be the name suffix and #barath an abstract noun from BAR- 'uplift, save, rescue' (orignial meaning 'raise'), MBAR- 'dwell, inhabit' or maybe from BARATH-. A third alternative would be ómataima-extended #bara- from BAR- and #thil. Compare also Tol Varad 'the Defended Isle' below.

• Looking at the names Barithil, Berithil, Baranor and Bergil one can see a striking pattern. The second elements are apparently Ithil 'moon' (THIL-, Silm.sil-), Anor 'sun' (ANÁR-) and gil 'star' (Rgeo:73). The first element may be #bar- < BAR- or MBAR- as mentioned, once #ber- with i-affection   The meanings thus seem to be *'Moon-protector', *'Sun-protector', *'Star-protector' or perhaps *'Moon-home', *'Sun-home', *'Star-home'. This would probably rather elliptically refer to the parts of Gondor or its former kings than to an actual protection of the astronomical objects. According to the outline in The Silmarillion and The Lord of the Rings the two founders of Gondor are Anárion reigning in Minas Anor 'Tower of the Sun' in the land of Anórien *'land of the sun'; and Ithildur reigning in Minas Ithil 'Tower of the Moon' in Ithilien 'land of the moon' (RC:233). In between lies Osgiliath 'Fortress of the Stars'. Tolkien calls the name Anórien 'heraldic' rather than climatic, and related to the heraldic names of Elendil's sons (RC:509). Minas Tirith [Anor], Minas Morgol [Ithil] as well as Ithilien, Anarion (Anárion in WR:243-244) and Osgiliath already appear on the maps in TI:308-309. Compare also Orendil, Ithildor and the river Ithilduin above.

• Duilas looks as if it contains dui- from DUI-, yielding words for 'river' and lenited glas 'joy' as in the name Borlas (GÁLAS-). Another possibility might be initial lenition of the root TUY- 'spring, sprout' (cf. G. duil, duilir 'spring' (GL:31)) and las 'leaf', thus *'Spring-leaf' (cf. Finduilas above).

• Garathon may be related to GARAT-, whence N. garth 'fortress', although it is probably untinterpretable as well.

• Targon is similar to Turgon, but most probably with initial tar- from TĀ-, TAƷ- 'high, lofty, noble'.

• Ramloth seems to mean *'wall-flower' (which is also a known group of flower species) with N. rhamb, rham 'wall' (RAMBĀ-), later S. ram (Silm.index) and N. lhoth 'flower' (LOT(H)-), later S. loth 'flower' (Silm.index), 'inflorescence, a head of small flowers' (VT42:18).

• For Thalion 'hero, dauntless man' see STÁLAG-.

• Gwinhir is probably *'young master'. Compare GWIN- 'new, fresh', Q. winya, N. bîn 'new, fresh, young' (VT45:16), later √WIN- 'young' with S. gwein 'young', gwîn 'youth' (PE17:191, VT47:26), S. gwinig *'baby, little young one' (VT48:6). Thus #gwin- 'young' and N. hîr 'master' (KHER-).

• Asgil seems of obscure origin, but may contain gil 'star' and #as- related to AS- 'warmth' (VT43:18). Golamir might be related to ÑGOL-  'wise, wisdom, be wise' and -mir might be N. mîr 'jewel, precious thing, treasure' (MIR-), also occurring in names like Boromir, Faramir.

• Inram, Benrodir and Nosdiligand all appear as chiefs of troops from various corners of Gondor marching into Minas Tirith and all of their names seem uninterpretable. Benrodir might contain #rodir *'nobleman' < rau-, ro- < √AR 'good, excellent, noble' (PE17:147) + †dîr 'man' surviving chiefly in proper names (DER-), but has an obscure initial element. The change of Nosdiligand to Northiligand leads to an extrapolation of #tiligand (differently mutated in the two names), similar to N. talagand 'harper' (ÑGAN-, ÑGANAD-), while #nos-, #nor- may relate to NŌ- 'beget' with N. noss 'clan, family, 'house'', nûr 'race'; but nothing more specific can be said.

• Dirgon seems to consist out of N. dîr 'man' and gond, gonn 'stone' (GOND-, Silm.index). If so, this should probably signify *'man of Gondor', compare also Bered Ondrath below. In his later name Hirgon the element dîr is exchanged with hîr 'master' (KHER-). On the other hand we know the lenited suffix -gon 'valour' (KAN-) appearing in personal names, thus Dirgon may also be interpreted as *'valiant man' and Hirgon as *'valiant lord'.
  So overall one has point out here once more that personal names of that kind need not to have a coherent meaning, moreover they need not to be fully Sindarin or Noldorin at all. See also the interpretations given in [8].

● Ramas Coren >> Ramas Ephel (WR:288,324) < N. rhamb, rham 'wall' (RAMBĀ-), later S. ram (Silm.index) + collective ending -as; coren *'circular' or *'encirling' < *korina < KOR- 'round'
● Corramas (WR:340)
These are earlier names of Rammas Echor or simply Rammas, an outer wall encircling Minas Tirith.
For Ephel 'outer fence' see Hebel Dúath above. Corramas *'round-wall' is obviously directly formed from KOR-, cor-.

● Othram 'City Wall' (WR:288) < N. ost 'city, town with wall around' (OS-) + N. rhamb, rham 'wall' (RAMBĀ-), later S. ram (Silm.index)

● Lonnath Ernin (WR:294)
This is an earlier name of Harlond, a haven south of Minas Tirith.
In the widely known Hiswelóke's Sindarin dictionary (2.0) a plausible interpretation of ernin as being the plural of arnen *'royal' < *arnina is proposed (cf. ƷAR-). Lonnath must be the collective plural of lond 'haven', in The Etymologies N. lhonn means 'narrow pass, strait, pass' (LOD-), but later S. lond, lonn 'haven' < LON (VT42:10).
The problematic point (also discussed there) is that there are also nearby hills called Emyn Arnen, where arnen would stand in the singular qualifying pl. emyn < amon 'hill' (AM2-). This fact, according to Tolkien's analysis of this name in VT42:17, means that arnen cannot be a Sindarin adjective, but instead may be wrongly used ar- 'beside' (in this meaning Quenya, not Sindarin) + nen 'water', thus 'hills beside the water'. But then Lonnath Ernin would not make much sense.
Perhaps the external development of these forms can be summarized as following: At first Tolkien thinks of arnen *'royal', pl. ernin and calls the haven by that name, but then abandons it, calling the haven Harlond and the nearby hills the 'Hills of Haramon' (see below). Later he decides that arnen should be an obscure word and changes the name of the hills to Emyn Arnen.

● Calledin (WR:301)
● Calembrith (WR:307)
These are other two names for the green lawn below Amon Hen, later to be called Parth Galen (LotRII, ch.10), earlier Calembel (and other forms - see above); Calembel also occurs beside Calledin.
Calledin seems to be formed with the element cal- from KAL- 'shine', but here it is probably meant to have the same developed sense 'bright-coloured = green' as N. calen. The second element looks like ledin 'fields', the plural of l(h)ad 'plain, valley', as supposedly in Palath-ledin 'Gladden Field[s]' (TI:114) (see above) - thus *'green plains/fields'.
Calembrith is apparently formed with calen 'green' and brith 'gravel' (BIRÍT-, Silm.index). This has perhaps either to be taken less literal as 'shore' or maybe it refers to grass growing through pebbles? Else Tolkien may have intended a reinterpretation of brith.

● Tol Varad 'the Defended Isle' >> Men Falros (WR:326)
● Cairros >> Andros >> Cair Andros (WR:340)
These are earlier names of Cair Andros 'Ship of Long-foam' (RC:544), a fortified island on the Anduin not far north of Osgiliath, so named for the isle was shaped like a great ship, with a high prow pointing north, against which the white foam of Anduin broke on sharp rocks (LotR App.A).
Tol Varad is obviously formed with N. toll 'island' (TOL2-), later S. tol 'isle' (Silm.index) and the lenited adjective barad < *baratā from BAR-, which yields Q. varna 'safe, protected, secure', N. berio 'to protect'. The adjectival ending *-tā occurs for instance in Q. pasta 'smooth' (PATH-), titta 'little, tiny' (TIT-) and so on. But then this word clashes with N. barad 'doomed' (MBARAT-) and barad 'tower, fortress' (BARAT-). Tolkien might have transcribed -ð by -d here, compare the commentary that had may stand for S. -hadh (VT42:20). If so, we would deal with the more frequent adjectival *-dā, as e.g. in tundā > N. tond, tonn 'tall' (TUN-) and with #barad = *baradh.
In Men Falros the former men perhaps means *'place', being cognate to Q. men 'place, spot' (MEN-), not otherwise attested in Noldori/Sindarin with this meaning (but see later Men-i-Naugrim 'the Dwarf Road' (UT:281)). Or perhaps men is in fact a derivative of *mindā < MINI- 'stand alone, stick up' meaning *'prominent tower/fortress'. Falros appears to consist out of PHAL- 'foam' and ROS1- 'distil, drip', ros 'foam, spindrift, spray' in The Silmarillion. Thus one may tentatively suspect either *'place/region of splashing foam' or *'tower/fortress of splashing foam'.
Cairros is apparently *'ship-foam' with cair 'ship' (KIR-), ceir in The Etymologies; and ros 'foam, spindrift, spray' (Silm.index).
As it is evident from the translation of the final name, Andros must signify 'long-foam' from N. and, ann 'long' (ÁNAD-, ANDA-) and ros.

● Fen Fornen, Fenn Forn 'the Closed Door', Fenn uiforn 'the Ever Closed' (WR:338,341)
● Uidavnen, davnan (WR:341)
These are earlier names of Fen Hollen 'Closed Door', see the description:
Turning westward they came at length to a door in the rearward wall of the sixth circle. Fen Hollen it was called, for it was kept ever shut save at times of funeral, and only the Lord of the City might use that way, or those who bore the token of the tombs and tended the houses of the dead. (LotRV, ch.4; RC:550)
The Etymologies give N. fend, fenn 'threshold' (PHEN-), but forn is the adjective 'right or north' (PHOR-), not 'closed'. So it seems that Tolkien planned a revision of roots or created a new one (homophone PHOR-, *SPOR-?) with the meaning 'close'. The form uiforn is then formed with the prefix ui- from OY- 'ever, eternal'.
The adjectives #tavnen < uidavnen *'ever-closed' and #tavnen on the other hand seem to be derived from TAP- 'stop', whence Q. tape 'he stops, blocks' (VT46:17), tampa 'stopper'. One can thus suspect *tapninā > *taphnina > tavnen or *tapnanā > *taphnana > tavnan. Compare lepn- > levnui (VT42:26)

● Bered Ondrath (WR:340)
This isolated name does not occur later and describes the guard-towers upon the entrance of the causeway, i.e. the causeway from Rammas Echor to Gondor. See the description of Rammas Echor in LotRV, ch.1:
At its furthest point from the Great Gate of the City, north-eastward, the wall was four leagues distant, and there from a frowning bank it overlooked the long flats beside the river, and men had made it high and strong; for at that point, upon a walled causeway, the road came in from the fords and bridges of Osgiliath and passed through a guarded gate between embattled towers.
Bered must be the plural of N. barad 'tower, fortress' (BARAT-), compare the commentary:
The 'correct' plural of onod was enyd, or general plural onodrim; though ened might be a form used in Gondor (Let:168)
The same seems to happen in the case of barad, which is after the fashion of Gondor umlauted to bered rather then *beraid or *bereid. In the Noldorin of The Etymologies, however, such pluralization was common, see [3].
Ondrath is here probably lenited *Gondrath in genitival position, with gond, gonn 'stone' (GOND-, Silm.index) + rath 'street (in a city)' from rath- 'climb' (RC:523, UT:255); thus *'stone-road, causeway' or perhaps elliptically *'Gondor-road'.

● Taur Rimmon (WR:350) < N. taur 'great wood, forest' (TÁWAR-), Rimmon is the name of a beacon-hill of pre-Númenórean origin (LotR App.F)
The beacon of Min-Rimmon stood on a tall hill amidst this forest, once mentioned by this name.

● fornest, Anfornest >> Forannest (WR:353-354,357)
● nest 'heart, core' (WR:357)
● nesta, nethra (ibid.)
The name Forannest appears on isolated notes and is once described as being the north gate in the Rammas, the name was not used later.
With the help of nest one can analyze fornest as N. #for- 'north' (PHOR-) or forn 'right or north' + nest 'heart, core' and Anfornest is put together with #ann 'gate' from AD- 'entrance' (compare Annerchin, Annerchion above; also Early Noldorin ann 'door' (PE13:137)) and fornest.
Forannest shows a different order of these elements - #for(n)- 'north(ern)' + #ann 'gate' + nest 'heart, core'.
The word nest 'heart, core' must be derived from NÉD-, whence N. enedh 'core, center', nedh- 'mid-'. For such a phonological development compare N. gwest 'oath' < wed-tā from WED- 'bind'. Thus fornest may signify *'centre north' in contrast with 'north-west' or 'north-east'. Due to the final vowels in nesta, nethra, they might be verbs *'centre, place in the centre'. While nesta may also be a primitive form of nest, nethra seems to show the late development -str- > -thr- as in ost-rond > othrond 'fortress, city in underground caves' (OS-).

● Haramon *'southern hill' (WR:359,363,434-435) < #har- 'south' (KHYAR-) + N. amon 'hill' (AM2-)
This is an earlier name of Emyn Arnen, see also Lonnath Ernin above. The name refers to several hills (the Hills of Haramon) despite the singular form.

● Haradoth >> Haradhoth (WR:365) < N. harad 'south' (KHYAR-) + hoth 'host, crowd, frequent in people-names as Glamhoth (KHOTH-)
For the loss of h compare Lossoth 'the Snowmen' (LotR App.A), Loss(h)oth (PE17:39), apparently formed with S. loss 'snow' (VT42:18); and also earlier Rochiroth, Rohiroth above.

● Berin a Nestad 'the Houses of Healing' >> Bair Nestedriu (WR:379-380)
● Bair Nestad (WR:380)
● {Edeb na Nestad} (ibid.)
These are Elvish names of the Houses of Healing in Minas Tirith, not used in The Lord of the Rings.
The Etymologies give bár 'home' (MBAR-, VT45:33), the Silmarillion index cites S. bar 'dwelling' and mentions that the old form mbár meant the 'home' both of persons and of peoples. Bar is pluralized to berin with the suffix -in in the first name, also seen in nauglin 'dwarves' (NAUK-) and probably ledin 'fields' in Palath-ledin, Calledin above.
Nestad looks like a gerund of a verb #nesta- and as exactly the same form occurred earlier (fornest, Anfornest above), a likewise derivation from NÉD- is worth a thought. This stem is related to ÉNED- 'centre' and also to NĒ- (NĔ-) (VT45:38), whence ndē-, ndĕ- 'in, inside', ne-stak- 'insert, thrust in, sting' > N. nestegi 'insert, thrust in'. If it is not too much of a lively imagination, the meaning 'heal' could have been developed from 'insert' in the sense of 'insert/take medicine'.
However, also probable seems a derivation from NETH- 'young' via *neth-tā > *nesta-. Compare host 'gross' < *khoth-tā from KHOTH- and perhaps similarly N. gost 'dread, terror', gosta- 'fear exceedingly', Gostir 'dread glance' from GOS-, GOTH-. On the other hand, th-t > th (þ) is in fact well-attested in several examples, e.g. ON pattha > N. path 'smooth' (PATH-).
Far more later NETH- is once glossed 'fresh, lively, merry' with Telerin nēþa 'gay, lively, girlish' (VT47:32-33). If similar connotations have already occurred earlier, 'fresh, lively' might have evolved into *'healthy, free from sickness' and hence *nesta- 'make healthy, heal, cure'. But of course, we may also be dealing here with an otherwise unattested root *NES-.
Berin a Nestad is formed with the genitive marker a, probably a shortened form of an - compare Rath a Chalardain, Ernil a Pheriannath above.
Bair Nestedriu is then apparently formed with i-affected plural of bar > bair (PE17:164). Nestedriu is most probably a misreading for *nestedrin, plural of the adjective *nestadren from the gerund nestad. Compare N. forodren 'northern' beside forod 'north' and N. haradren 'southern' beside harad 'south' (PHOR-, KHYAR-).
In Bair Nestad the gerund nestad is put into trailing position as uninflected genitive.
Edeb na Nestad transparently contains N. edeb, pl. of adab 'building, house'. Compare the pluralization barad > bered in Bered Ondrath above. The genitive marker is now na, as also in various other attested forms, cf. Henlo n'Annun above.

● asea aranaite >> asea aranion 'kingsfoil' (WR:394)
● athelas (ibid.)
For the formation of aranaite *'royal' compare maʒiti 'handy, skilled' > Q. maite (MAƷ-), kiryaite beside kiryăva (PE16:113) and Q. aran 'king' (WJ:369). Aranion is transparently the Quenya genitive plural of aran.
For the interpretation Tolkien gives the root √ATHA with Q. asya- 'to ease, assist, comfort, the noun asië 'ease, comfort', asëa 'as name of plant "athelas"', S. athae, ath(a)elas and the verb eitha- 'to ease, comfort' (PE17:148).

● tarakil 'Trotter' (WR:390,395)
● Tarakon >> Tarantar >> Telkontar (WR:395)
This is another set of earlier names for 'Trotter', obviously in Quenya. As Aragorn says:
But Trotter shall be the name of my house, if ever that be established; yet perhaps in the same high tongue it shall not sound so ill, and tarakil I will be and all the heirs of my body. (WR:390)
Later, Tarantar is translated as 'Trotter' and Telcontar as 'Strider' (SD:121).
Both tarakil and Tarakon seem to contain #tarak- with two different name-formative suffixes -il and -on attached. A sufficient explanation is problematic, since the only suitable root TARÁK- means 'horn (of animals)' > Q. tarka 'horn'. However, Goldogrin had the word tarc 'root' (Q. tarka, GL:69) beside târ 'a horn'. And since we know TÈLEK- 'stalk, stem, leg' yielding telch 'stem' in Noldorin and telko 'leg' in Qenya, we could suppose that 'leg' and 'root' may also share a common origin *tarak-.
Another root in question is √TĂR 'stand, intransitive' (PE17:186) - the meaning 'stand' could have likewise yielded 'leg'. Compare TAL- 'foot' and its extension TALAM- 'floor, base, ground'.
On the other hand in the Qenya Lexicon the roots TARA and TARAMA 'to batter, thud, beat' are given with derivatives like Q. taru 'horn', tartan 'hammer' (QL:89). The Gnomish equivalents are mentioned as DAR-, D(A)RAM-, DAM and in The Etymologies we find TAM- 'knock', NDAM- 'hammer, beat', all probably being onomatopoetic, describing a beating sound. Hence, a related root *TARA(K)- could have yielded 'trot, tread' also by the time of The Etymologies. In this context compare Q. pata- 'rap, tap (of feet)', patake 'clatter', patakatapaka 'rat-a-tat' with later ap-pata 'walk behind, on a track or path' (WJ:387).
Tarantar could be from *TARAM- with the suffix -tar being either name-formative as -il, -on or a reduplication of the initial element. Alternatively, we may be dealing here with a verb #taranta- *'trot' and agentive formation, as for instance Q. mahta- 'wield a weapon, fight', mahtar 'warrior' (MAK-).
Telkontar seems to be formed with Q. telko 'leg' as already mentioned; and #-ntar may be playing here the same role as in Tarantar, i.e. either indicating that a name was formed from telko (perhaps via the verb #telkonta- *'trot' or *'stride') or forming the compound *'leg-tread, leg-trot'.

● Barad Amroth 'Castle Amroth' (WR:409,423,424) < N. barad 'tower, fortress' (BARAT-)
This was a passing substitution for Dol Amroth.

● Lothland (WR:426)
● Lostladen (WR:435)
These are two variants for the name of the desert south of Mordor. Earlier it was called Lothlann 'wide and empty' (TI:313), see above.
Lothland is apparently a mere variation - both Lothland and Lothlann already appear in The Etymologies under LAD- with N. lhand, lhann 'wide'.
Lostladen is obviously formed again with l(h)ost 'empty' (LUS-), but a different adjective l(h)aden 'open, cleared' (LAT-). For the preservation of medial -stl- in contrast to -stl- > -thl- in Lothland, Lothlann compare N. mistrad 'straying, error' (MIS-), ostrad 'a street' (RAT-).

● Caerost on Kiril (WR:437)
● Tarnost (ibid.)
These are names of two cities in Gondor.
Caerost preceded Calembel (see Calledin, Calembrith above). It obviously contains N. ost 'city, town with wall around' (OS-). Caer is attested in Noldorin as the numeral 'ten' (KAYAN-, KAYAR-), thus *'ten-town'? Else it might be a derivative of KAY- 'lie down', whence N. caew 'lair, resting-place', thus *'lying/resting town'.
Tarnost was a pencilled addition to the map, located on the river Ringlo. The initial element #tarn- is the same as in earlier Tarn Felin (TI:424), see above. Taking Early Qenya tarna 'crossing, passage, ford' into account, as well as the town's position on a river, it might indeed be the *'crossing-town' or *'ford-town'. Compare also THAR- 'across, beyond', thar- 'athwart, across' (Silm.index), whence another city name is derived: Tharbad < thara-pata 'cross-way'. See also Bronwe athan Harthad 'Endurance beyond Hope' (SD:62), where athan may be a misreading for *athar 'beyond' < THAR-. In Quenya the preposition tar 'beyond' occurs once (LR:72), rather then *sar from THAR-.
It is hardly more than a guess, but maybe there was a parallel stem *TAR-, likewise meaning something like 'across, beyond', yielding the Qenya preposition tar, but the Noldorin/Sindarin noun #tarn *'crossing, ford'. Note also the frequent occurrence of English ford, German Furt, Dutch voorde in 'real' place names.
But we also find a directly attested root √TĂR 'stand, intransitive' (PE17:186) and tarn *'standing' could be an adjectival derivation of it. Hence Tarnost *'standing town' in symmetry with Caerost *'lying/resting town'.

● Gorgoroth >> Narch Udûn (WR:438)
These were earlier names of the vale behind Morannon, called Udûn in The Lord of the Rings (LotRVI, ch.2).
The first name is simply N. Gorgoroth 'deadly fear' (ÑGOROTH-).
Narch probably means here *'cleft' from NÁRAK- 'tear, rend (tr. and intr.)', compare Narchost 'bitter-biting fort' (RC:601). N. Udun  < Utubnu (TUB-) was originally Melkor's northern fortress in The Silmarillion, compare also the entry tum of its index. Although not directly translated, it must have connotations with great deepness.

Observations

Much less is written by Tolkien himself and published about the 'Celtic branch' of Eldarin languages, i.e. Noldorin, Sindarin and Ilkorin, than about Quenya, so that every bit of information is very helpful. Therefore a short final summary and conclusion seems necessary.
 

mutation and assimilation

As lenition is the main driving force of these languages both in phonological development and grammar, it is worthwhile to take a look at the names under this viewpoint. Surprisingly there is little of it in many of the words discussed above. For example,

    ● Rathcarn 'Redway' < rath 'way' + #carn *'red'
    ● Criscarn 'Red Pass' < criss 'cleft' + #carn *'red'
    ● Narosîr 'Redway' < sîr 'river'
    ● Rhascaron *'Red Horn' < rhas 'horn' + caron *'red'
    ● Erceleb  < er- + celeb 'silver'
    ● Pensarn *'stony slope' < pend 'declivity' + sarn 'stone'
    ● Tum Dincelon 'Dimrill Dale' < #dim- *'dark, gloomy' + celon 'river'
    ● Eredhithui 'Misty Mountains' < orod 'mountain', pl. ered + #hithui 'misty'
    ● Llawhen *'ears-eye' < llaw 'ears' (dual) + hen 'eye'

show no lenition of the second element at all, Rathcarn has been even changed from Rathgarn. This is surprising, as we know words like basgorn 'loaf' < bast-gorn < *bast-corn 'round bread' (MBAS-, KOR-, once glossed Ilkorin and once Noldorin), where in the contact s-c the second word is lenited, although the combination -sc- seems to be permitted both in Noldorin and Ilkorin when it comes from the same stem, compare N. asgar, ascar, Ilk. ascar 'violent, rushing, impetuous' (SKAR-).
Even more surprising is Erceleb, a unique example of an unchanged contact of r-c, although Erchamui 'One-handed' (KAB-) and numerous other words show that a stop after r would become a spirant or at least undergo lenition.
Pensarn has unchanged contact n-s, while we can see from the name Arassuil (LotR App.A) < *aran-suil 'king-greeting' that n-s assimilates to -ss- in later Sindarin; although it may be here due to the medial simplification of *pendsarn > pensarn. Still, one would have at least expected *Penharn, as in Calenhir, Tolharn. Note that according to a conception of the later Northern Sindarin dialect s is unlenited initially (PE17:134) and it was also unlenited in Goldogrin and Early Noldorin (GL:7, PE13:121) - perhaps the same applies at that time to Ilkorin?
In Dincelon the word celon is not lenited; instead we are perhaps observing *dim- assimilating to the following c-. However, in examples like

    ● Carnbeleg < carn 'red' + beleg 'mighty'
    ● Mornvenniath 'Black Mountains' < morn 'black' + ment 'point' + iâth 'fence'
    ● Calenbel, Calen-Bel < calen 'green' + pel(?) 'fenced field'

we see even a lack of assimilation, with the clusters -rnb- and -rnv- allowed medially, as well as -nb-, although Calenbel or Calen-Bel was changed to Calembel, Cálembel and later Calembrith can be found.

Lenition occurs in the following words:

    ● Beleghir *'mighty river' < beleg 'mighty' + sîr 'river'
    ● Calenhir *'bright/green river' < calen 'bright-coloured, green' + sîr 'river'
    ● Tolharn *'stony island' < toll 'island' + sarn 'stone as a material; or as adj.'
    ● Nenvithim 'Hoardales' < nan 'plain, valley' (pl. *nen) + #mithim *'hoar, grey'
    ● Hithdilias 'Misty Mountains' < hith 'mist' + #tilias 'mountains'
    ● Eredwethion ‘Mountains of Shadow’< orod 'mountain', pl. ered + gwath 'shade' + -ion
    ● Eredvyrn 'Black Mountains' < orod 'mountain', pl. ered + morn 'black', pl. myrn
    ● Egladil 'elven-point' < egla 'elf' + till 'point'
    ● Calendil 'Green Spit, Green-tine' < calen 'bright-coloured, green' + #-til 'point'
    ● Celebdil 'Silverhorn' < celeb 'silver' + till 'point'
    ● Duin Morghul *'river of sorcery' < mor- 'black' + gûl 'magic'
    ● Dol Dúghul 'Hill of Sorcery' < dû 'nightfall, late evening' or 'night, dead of night' + gûl 'magic'
    ● Ered Orgoroth 'Mountains of Terror' < orod 'mountain', pl. ered + gorgoroth 'terror'
    ● Angrobel 'Isengard' < #angr- 'of iron' + gobel 'walled house or village, 'town''
    ● Fornobel 'Norbury' < forn 'right or north' + gobel 'walled house or village, 'town''
    ● Tor-dilluin *'Mighty Bluehorn' < taur 'vast, mighty, overwhelming, awful, high, sublime' + #til(l)-luin 'horn-blue'
    ● Nargalad, Anngalad, Carangalad *'red light' or *'light of the west' < calad 'light'
    ● Galbedirs < #gal- 'tree' + #pedir 'speaker'

One has to point out Beleghir and Calenhir in contrast with Narosîr. The usage of the circumflex for the second unlenited element in Narosîr is noticeable - it is usually used in monosyllables and the other compounds with lenited sîr shorten their vowel. Perhaps this is indeed a different (more loose) kind of compound *Naro-sîr and therefore without lenition.
The lenition of g- to the spirant gh in Morghul, Dúghul ignoring primitive ñg- is also different to later Sindarin, where we find i ngaurhoth *'the wolf-host' (LotRII ch.4, ÑGAW-), di-nguruthos 'beneath-death-horror' (LotRIV ch.10, Rgeo:72, ÑGOROTH-, ÑGUR-); also earlier di-ngorgoros (or di-ngorgoroth, reading uncertain) in The Etymologies (VT45:37). Such a special treatment of former nasalized stops goes in fact all the way back to the Gnomish Lexicon (e.g. golda > i·Ngolda 'the gnome' (GL:8)).
Lenition g > gh may be either an analogical development - historical forms would tend to become forgotten by the time of the Third Age, compare dor 'land' > i·nnor, indor 'the land', analogical i·dhor (PE13:161). The back spirant surely strengthens the effect of -ghul applied to dark magic.
But in Angrobel, Fornobel, Ered Orgoroth initial g- is lenited to zero (gorgoroth may have also had initial ñg-, but The Etymologies give primitive gor-ngoroth). Probably there had been lenited *ʒobel, *ʒorgoroth at an intermediate stage before the spirant fell away, although remaining in monosyllabic elements (as Mor-ghul, Dú-ghul) - otherwise they would have become unrecognizable. Or gh only remained after r or a vowel and fell away after other consonants or clusters. Compare the later Sindarin development nāba-grota 'hollow' + 'excavation, underground dwelling' > nǭv-ʒrot > novrod (WJ:414).

The forms with hoth 'host' show three different results - lenition, no lenition and the dropping of h. Perhaps this has to be understood as a suffix rather than part of a compound:

    ● Rochirchoth
    ● Haradhoth, Rochirhoth
    ● Haradoth, Ro(c)hiroth

There are some examples of trailing and lenited adjectives:

    ● Caradras dilthen 'little Redway', probably < *tilthen 'little'
    ● Tol Galen *'Green Isle' < calen 'bright-coloured, green'
    ● Tol Varad 'the Defended Isle' < #barad 'defended'
    ● davnan < *tavnan 'closed', from the context: Fenn davnan 'the Closed Door'

But there are about twice as much without lenition:

    ● Cris Caron 'Red Pass' < caron *'red'
    ● Ered Myrn < morn 'black', pl. myrn (but changed to Eredvyrn)
    ● compounded Eredhithui < hithui *'misty'
    ● Ennyn Dûr < dûr 'dark, sombre'
    ● Nelig Morn >> Nelig Myrn, also Naglath Morn *'black teeth' < morn, pl. myrn 'black'
    ● Henneth Malthen *'golden window' < malthen 'of gold, golden'
    ● Ramas Coren *'circular/encircling wall' < coren *circular' or *'encircling'
    ● perhaps Taur Toralt with toralt *'mighty and large'

The interpretation of Arad Dain *'High Pass' is too uncertain to work with. Minas-berel may also contain an unlenited adjective berel *'valiant'.

a-affection

Adjectival endings like -imā or -(r)inā are found in many Elvish words and Tolkien was not always sure about their fate in the 'Celtic branch'. In Noldorin and Sindarin we see a shift i > e caused by final -ā [4]. In the sources discussed three forms look like candidates for the lack of a-affection, but all of them may be Ilkorin which does not show a-affection [1]:

    ● Celebrin 'of silver' < *kyelep-rinā
    ● Palathrin 'Gladden river' < palath 'iris, gladden' + -rinā
    ● Nenvithim 'Hoardales' < #mithim 'hoar' < *mithimā
    ● Neleg Thilim 'the Gleaming Tooth' < *thilimā '(white) gleaming'

A-affection in an isolated word is seen in:

    ● coren *'circular/enclosing' < *korinā

Three names (if correctly analyzed) show forms with a-affection medially in a compound:

    ● #borthen *'steadful' (from Borthendor) < *bortinā
    ● Calenhir *'green/bright river', Calenardhon *'green realm' < *kalinā (also Kalenarda)

One name shows probably no a-affection medially:

    ● melinon *'yellow one' < malina < *malinā

With adjectives in such a position it probably depends on the time of compound formation - whether before or after the shift i > e. Thus Melthinorn *'Gold-tree' (a name of Laurelin; SMAL-) beside N. malthen 'of gold' is an ancient name from the First Age, whereas Calenardhon, Calenhir, Borthendor are regions named in the Third Age.
 

transcription

At that time Tolkien considered the spelling of [k] with the letter k instead of c also for the 'Celtic branch' and not only for Quenya as before, e.g.:

    ● Kelebras 'Silverhorn'
    ● Kelegorn *'swift-impetuous' [considered as the name of Aragorn's father]
    ● Kelufain *'white source'
    ● Keleborn *'silver-tree'
    ● Kerin-muil *'hidden enclosure'(?)
    ● Kirith Gorgor 'the Dreadful Pass'
 

Note on the endings -on, -ion

The endings -on and -ion are quite frequent in names, but their role is not always clear. So I will try to treat them more carefully.

Already in the earliest Celtic-style language Goldogrin both -on and -ion are adjectival endings (beside -n, -in), e.g. argulthion 'equal, equivalent' (GL:20), gwedhwion 'looped, bending' (GL:46), martion, mart 'fated, doomed, fey' (GL:56), taithion, godaithion 'educated' (GL:68) or falon, falin 'naked' (GL:33), helon 'frozen' (GL:48), hebon 'bound - also bounded, surrounded' (ibid.), malon 'yellow' (GL:56), melon, meltha 'dear, beloved' (GL.57).
At the same time -ion is the genitive plural of consonantal nouns, e.g. glôr 'gold' >  glorion; and -on is the genitive singular of nouns ending in -a/-u, as coma 'disease' > comon, culu 'gold' > culon (GL:12-14).

By the time of The Etymologies Noldorin, Ilkorin and Doriathrin are the languages of the Celtic branch. Both -on and -ion occur as agentive suffixes, as in  #faron 'hunter' (SPAR-), Dúrion/durion 'a Dark-elf' (DOƷ-, DÔ-, MOR-) or else form names as Mirion 'ordinary N name of the Silevril (Silmarilli)' (MIR-), Gelion 'merry singer', also a river (GYEL-). Probably associated with this is the patronymic suffix -ion < YŌ, YON- 'son'. Tilion 'hyrned' or ‘the Horned’ (TIL-), Brithon 'pebbly' (BIRÍT-) and Erchamion 'one-handed' (beside Erchamui, Ermabuin, Ilk. Ermab(r)in (MAP-, LR:427,146,405)) look like adjectival forms but are also names at the same time. N. Erchamron and later S. Erchamon 'one-hand Man' (VT47:7) are no adjectives. Purely adjectival seem to be Ilk. gelion 'bright' from GAL-, tovon 'lowlying, deep, low' < tubnā (TUB-).
At the same time Ilkorin shows the genitival inflection sg. -a, pl. -ion as in Dor-thonion 'Land of Pines', Torthurnion 'King of Eagles' [1]. The genitive pl. -ion is probably also found in Noldorin in Eredwethion ‘Mountains of Shadow’ (TI:345, WATH-). This is explicitly Noldorin, the Ilkorin variant being Urthin Gwethion. However, another translation is 'Shadowy Mountains' (LR:447) hinting at an adjective #gwethion 'shadowy'. Also Duil Rewinion 'Hills of the Hunters' (LR:286).
But that is not all yet - the simple addition of -ion seems to denote a region: Dor. Regornion 'Hollin' (ÉREK-) from regorn 'holly'.
The suffix -on (and -ion for i-stems) is also often augmentative: N. annon 'great gate' (AD-), later S. (g)aearon 'the Great Ocean' (Rgeo:72-73, PE17:27) < aear 'sea', Sirion 'the Great River' < sîr 'river' (Silm.index).
See also [2].

The close association between patronymic -ion and gen. pl. -ion that also occurs in Quenya and is given account in the Early Qenya Grammar:
-ion, old patronymic ending, which has appearance of being a genitive plural and hence is often formed from -li form of vocalic nouns, as Noldolion (pl. noldoliondi) 'descendant of the Gnomes' (PE14:45).
In Noldorin/Ilkorin names -ion is indeed of a manifold ambiguity. So how should it properly be: Dufinnion *'dark-hair-person' or adj. *'dark-haired'; Annerchion *'gate of goblins' or *'goblinish gate'; Torfirion *'high-man + name-suffix' or maybe *'high-men-place', Andon *'long one' or adj. *'very long', Amarthon adj. *'doomed', *'greatly doomed' or *'doom + name-suffix', Duil Rewinion *'hills of hunters',*'hunting hills', *'hills of the hunting region', Eredwethion 'Mountains of Shadow' (gwath, gen. pl. #gwethion), 'Shadowy mountains' (adj. #gwethion) or even *'mountains of the shadow-region'?
It will not be surprising that Tolkien often adjusted the conceptions and interpretations when it came to explanations.

Thus according to one explanation Eregion 'Hollin' (cf. 'holly-region' in RC:772) < S. ereg 'holly' and Nanduhirion are said to add the regional ending -ion (PE17:42) which is the adjective iaun 'large, extensive, wide'. Related are older -ian(d) in Beleriand and pl. -iend, ien often used of a single varied region as Anórien, Ithilien. In older names it usually applied only to a large feature, as in Sirion 'the Great Stream' (PE17:42).
Another explanation gives -ion < -ı̯aun, from yānā < √YANA 'wide, large, extensive'; also S. iaun 'roomy, wide, extensive' (ibid.). It was applied especially to topog[raphical] features of large extent, especially long, wide rivers, long (and wide) ranges. So Sirion < siriānā; Eregion, Erydweithion, -ian. Compare essentially the same √YAN- 'vast, huge', untranslated S. iaun as part of Rhovanion 'Wilderland' (PE17:99).
But then Tolkien decides that Eryd-weithian(d) should = mountains of the region of the shadows. But then gweithian 'region of the shadows' should remain unlenited in genitival position. Hence Tolkien concludes: Better return to Eryd-wethrin, shadowy mountains (i.e. with an adjectival suffix). This is how the mountains appear in The Silmarillion - Ered Wethrin.
He further decides: Delete entirely yondo = 'son'! Very unsuitable (PE17:43) and comes up with a new etymology: √YŎNO 'wide, extensive', in regional names yonde, ionde > -ion, yon. Often associated with genitive plural [Also confused with -on, augmentative or male suffix]. So now Sirion is properly 'the Vale or lands about the River Síre' or 'the great stream'.
A yet different root is √YOD- 'fence, enclose' yields yonde 'any fairly extensive region with well-marked natural bounds (as mountains or rivers)' > -yonde, yon / ionde, ion frequent in regional names.

The genitive plural is just briefly mentioned here, but is explored in other notes. Thus Sindarin has gen. sg. -a, pl. -on called purely possessive (PE17:97). For the plural both endings may be combined with the ordinary pluralization via -i: lais galaðon or lais geledhion 'the leaves of trees', similarly glim maewion or glim maewia '(the) voices of gulls'. See also elenath 'the (host of all the) stars' > full genitive elenathon (PE17:24-25).
A remark reads: ion is ia < g[enitive] iōm, later n [?restored]. So presumably the plural marker -n is (for some reason - perhaps by Quenya influence?) restored after having been regularly lost -iōm > *-iōn > *-io > (unstressed?) -ia. (cf. VT42:14, VT47:24 and ai-lin- > N. oel 'pool, lake', pl. oelin < *ai-lini (AY-)).
At the end, however, Tolkien rethinks it: X DON'T have inflected genitive!

Other Sindarin samples with -on, -ion include:

Nan Gondresgion 'Stonewain Valley' < rasg 'a drag or any large, flat vehicle on wheels or rollers for hauling stone or other weighty material'; Q. Nand' Ondolunkava or Ondolunkanan(do) (PE17:28)
Dorwinion 'Young-land country' or 'Land of Gwinion' (Hobbit, PE17:54), cf. S. √WIN- 'young' (PE17:191) and earlier Ardwiniel (VT45:16)
Orgilion 'day of the Stars' (LotR App.D), earlier Argilion (PM:130)
Dorthonion 'Land of Pines' (Silm.index, PE17:81)
Loeg Ningloron 'Gladden Fields' or lit. 'Pools of the golden water-flowers' (UT:281,450)
Caras Galadhon 'City of the Trees' (LR:345,UT:425, LotRII, ch.7), originally Caras Galadon (cf. RC:311)

Also here different explanations can be found. According to one Galadon is actually a lenited adjective Caladon < calatāna (PE17:84). According to another Galad and especially the apparent genitive plural Galadon 'of trees' are not Sindarin (PE17:51), but rather Nandorin with galadā > galad (S. galadh). Although the spelling was changed to Galadhon with proper Sindarin dh, caras is assigned to Silvan speech in UT:257. So the whole name should be perhaps regarded at least as dialectical, if not full Nandorin. The Gladden Fields also lie to the east of the Misty Mountains and fit into this scheme.

From Let:347 we learn that Orbelain was a 'phonological' translation by the Noldor. So all week days, including Orgilion could be Quenya-influenced. Orbelain includes the reconstructed adjective belain *'of the Valar' (Q. Valanya), so maybe gilion is an adjective as well, influenced by the gen. pl. -ion in Quenya. Oraearon might also contain an -on adjective, corresponding to Q. Eärenya, or else S. (g)aearon 'ocean' (PE17:27).
This is also supported by the gloss of Dorthonion as S. Noldorized.

In Nan Gondresgion the position of Gondresgion as a qualifier as well as the Quenya version employing the possessive case suggest that the purely possessive genitive has been intended here.

On the other hand, the sole compound Dorwinion and its translation 'Young-land country' suggests an analysis dor-win- 'young-land' + -ion 'country'. In any case a genitive plural is not appropriate here. The river name Gwinion follows the pattern of Gelion, Tilion, Mirion and might itself be an adjective *'young'.

Referred sources and related articles:

[1] Ilkorin - a "lost tongue"? by Helge Fauskanger
[2] Compounds in the Noldorin of the Etymologies by Thorsten Renk
[3] Noldorin plurals in the Etymologies by Bertrand Bellet
[4] Vowel affection in Noldorin and Sindarin by Bertrand Bellet
[5] The Two Phonetic Values of ll in Elvish Sindarin in The Lord of the Rings by Carl Hostetter
[6] Northern Dialect of Sindarin by Ryszard Derdzinski
[7] Un arbre mystérieux : le lebethron by Didier Willis
[8] A Tolkien dictionary by Robert Ireland and Andreas Möhn

Roman Rausch (Aran) aranwe@mail.ru
Dec. 23^th, 2006

[update: Apr 3rd '07 ~ names from WR added, some additional comments added to the older entries]
[update: Oct 8th '07 ~ cross-references and interpretations from PE17 added, note on -on, -ion rewritten]