The Rivers and Beacon-hills of Gondor

The names of the Rivers

The essay begins with the lengthy excerpt and author's note given in UT:264-65 (and so not reproduced here). A few variances between the published text and the typescript are noteworthy: where the published text has Enedwaith the typescript reads Enedhwaith (this was an editorial change made in all excerpts from this essay containing the name in Unfinished Tales; cf. XII:328-29 n. 66); and where the published text has Ethraid Engrin, the typescript has Ethraid Engren (but note (Ered) Engrin, V.348 s.v. ANGĀ-, V.379 s.v. ÓROT-, and many other places beside). In addition, a sentence referring to the ancient port called Lond Daer Enedh was omitted before the last sentence of the author's note on UT:264; it reads: "It was the main entry for the Númenóreans in the War against Sauron (Second Age 1693-1701)" (cf. LR:1058; and UT:239, 261-65). Also, against the discussion of the approach to Tharbad that closes the first paragraph on UT:264, Tolkien provided the cross-reference"I 287,390".{10}

Following the passage ending at the top of UT:264, the essay continues with this etymological discussion, in reference to the name Glanduin:

glan: base (G)LAN, ‘rim, edge, border, boundary, limit'. This is seen in Q. lanya verb ‘bound, enclose, separate from, mark the limit of; lanwa ‘within bounds, limited, finite, (well-)defined'; landa ‘a boundary'; lane (lani-) ‘hem'; lantalka ‘boundary post or mark'; cf. also lanka ‘sharp edge (not of tools), sudden end', as e.g. a cliff-edge, or the clean edge of things made by hand or built, also used in transferred senses, as in kuivie-lankasse, literally ‘on the brink of life', of a perilous situation in which one is likely to fall into death.

It is debated whether gl- was an initial group in Common Eldarin or was a Telerin-Sindarin innovation (much extended in Sindarin). In this case, at any rate, the initial gl- is shared by Telerin and Sindarin and is found in all the derivatives in those languages (except in T. lanca, S. lane, the equivalents of Q. lanka): T. glana 'edge, rim';{11} glania- ‘to bound, limit'; glanna ‘limited, bounded'; glanda ‘a boundary': S. glân, ‘hem, border' (of textiles and other hand-made things), gland > glann ‘boundary'; glandagol ‘boundary mark';{12} gleina- ‘bound, enclose, limit'.{13}

Tolkien then comments: "The names of the Rivers give some trouble; they were made up in a hurry without sufficient consideration," before embarking upon a consideration of each name in turn. Significant portions of this section of the essay have been given in Unfinished Tales. Extended passages are not repeated here, but their places in the essay are indicated.

This is not on the map, but is given as the name of the short river flowing into the Isen from the west of Ered Nimrais in App. A, III 346.{14} It is, as would be expected in any name in the region not of Rohanese origin, of a form suitable to Sindarin; but it is not interpretable in Sindarin. It must be supposed to be of pre-Númenórean origin adapted to Sindarin.{15}

Of this entry, Christopher Tolkien notes: "On the absence of the name on the map—referring of course to my original map to The Lord of the Rings, which was replaced long after by the redrawing made to accompany Unfinished Tales—see UT:261-62, footnote."

Of the next entry, headed "Gwathlo (-ló)", Christopher Tolkien writes: "The long discussion arising from this name is found in UT:261-63, with the passage concerning the Púkel-men removed and cited in the section on the Drúedain, UT.383-84. In the latter passage the sentence ‘Maybe even in the days of the War of the Ring some of the Drú-folk lingered in the mountains of Andrast, the western outlier of the White Mountains' contains an editorial change: the original text has ‘the mountains of Angast (Long Cape)’,{16} and the form Angast occurs again more than once in the essay. This change was based on the form

Andrast communicated by my father to Pauline Baynes for inclusion, with other new names, on her decorated map of Middle-earth; see UT:261, footnote." A further editorial change may be noted: where the published text has Lefnui (UT:263, repeated in the extracted note on Púkel-men, UT.383) the typescript reads Levnui; cf. the entry for Levnui below.

An unused note against "the great promontory ... that formed the north arm of the Bay of Belfalas" (ibid.) reads: "Afterwards called still Drúwaith (Iaur) ‘(Old) Pukel-land', and its dark woods were little visited, nor considered as part of the realm of Gondor" Also, a sentence struck through by Tolkien, following "huge trees ... under which the boats of the adventurers crept silently up into the unknown land", reads: "It is said that some even on this first expedition came as far as the great fenlands before they returned, fearing to become bewildered in their mazes."

The discussion originally continued with the following etymological note, struck through at the same time as the deleted sentence:

So it was that the river was called in Sindarin Gwathlo (in Adunaic Agathurush) ‘the flood under shadow'. Gwath was a Sindarin word from a Common Eldarin base Wath or extended Wathar. It was much used; though the Quenya relative waþar, later vasar, was not in daily use.{17} The element -lo was also of Common Eldarin origin, derived from a base (s)log: in Common Eldarin sloga had been a word used for streams of a kind that were variable and liable to overflow their banks at seasons and cause floods when swollen by rains or melting snow; especially such as the Glanduin (described above) that had their sources in mountains and fell at first swiftly, but were halted in the lower lands and flats, *sloga became in Sindarin lhô; but was not in later times much used except in river or marsh names. The Quenya form would have been hloä.

This passage contains a note, also struck through, on the name Ringló, occurring after "Sindarin lhô", given in the discussion of that entry below.

The deleted passage was replaced with that given at UT:263 starting at "So the first name they gave to it was ‘River of Shadow', Gwath-hîr, Gwathir". It may be noted that the word lo in this passage was corrected on the typescript from lhô. A note on the name Ringló, omitted from the passage in Unfinished Tales, occurs after the words "Gwathlo, the shadowy river from the fens" For this note, and its development, see the entry for Ringló below. After this note, an etymological statement intervenes before the last full paragraph of the excerpt published in Unfinished Tales:

Gwath was a common Sindarin word for ‘shadow' or dim light—not for the shadows of actual objects or persons cast by sun or moon or other lights: these were called morchaint 'dark-shapes'.{18} It was derived from a Common Eldarin base WATH, and appeared also in S. gwathra- ‘overshadow, dim, veil, obscure'; gwathren (pl. gwethrin) ‘shadowy, dim'. Also related was auth ‘a dim shape, spectral or vague apparition, from *aw'tha. This was also found in Quenya auþa, ausa of similar sense; but the stem was otherwise only represented in Quenya by the extension waþar, vasar ‘a veil’, vasarya- ‘to veil'.

Lô was derived from Common Eldarin base LOG ‘wet (and soft), soaked, swampy, etc.' The form *loga produced S. lô and T. loga; and also, from *logna, S. loen, T. logna ‘soaking wet, swamped'. But the stem in Quenya, owing to sound-changes which caused its derivatives to clash with other words, was little represented except in the intensive formation oloiya- ‘to inundate, flood'; oloire a great flood'.

Against the words "owing to sound-changes which caused its derivatives to clash with other words" Tolkien added this note:

Thus the Quenya form of S. lô would have been *loa, identical with Q. loa < *lawa ‘year'; the form of S. loen, T. logna would have been *lóna identical with lóna ‘pool, mere' (from base LON seen also in londe ‘haven, S. land, lonn).

Though this was the first of the Rivers of Gondor it cannot be used for ‘first'. In Eldarin er was not used in counting in series: it meant ‘one, single, alone. erui is not the usual Sindarin for ‘single, alone: that was ereb (< erikwa; cf. Q. erinqua); but it has the very common adjectival ending -ui of Sindarin. The name must have been given because of the Rivers of Gondor it was the shortest and swiftest and was the only one without a tributary.

Against the words "the very common adjectival ending -ui of Sindarin" Tolkien added this note:

This was used as a general adjectival ending without specialized significance (as e.g. in lithui ‘of ash', or ‘ashen, ash-coloured, ashy, dusty'). It is of uncertain origin, but was probably derived from the Common Eldarin adjectival -ya, which when added to noun-stems ending in C.E. -o, -u would produce in Sindarin -ui. This being more distinctive was then transferred to other stems. The products of āya > oe, and of ăya, ĕya, ĭya > ei; ŏya, ŭya > æ, e were not preserved in Sindarin.{19} But -i, which could come from ēya, and from īya, remained also in (more limited) use; cf. Semi below. The transference is exemplified in the ordinals, which in Sindarin were formed with -ui from ‘fourth' onwards, though -ui was only historically correct in othui ‘seventh' and tolhui 'eighth'. ‘First' was in older and more literary Sindarin mein (Q. minya); later minui was substituted [deleted: in the colloquial language; ‘second' tadeg; ‘third' neleg]; but ‘fourth' cantui (canhui), ‘sixth' encui, enchui,{20} ‘ninth' nerthui [deleted: ‘tenth' caenui],{21} etc. On ‘fifth' see below under the name Lefnui.

Christopher Tolkien writes: "The statement about this name is given in the Index to Unfinished Tales, but with a misprint that has never been corrected: the Sindarin word meaning ‘pebble' is sarn, not sern." The opening sentence reads: "An adjectival formation from S. sarn ‘small stone, pebble (as described above), or a collective, the equivalent of Q. sarnie (sarniye) ‘shingle, pebble-bank." An unused sentence, occurring before "Its mouth was blocked with shingles" reads: "It was the only one of the five to fall into the delta of the Anduin."

This means simply ‘a flowing': cf. tirith ‘watching, guarding’ from the stem tir- ‘to watch'.

Christopher Tolkien writes: "The statement about this name is given in the Index to Unfinished Tales. On the erroneous marking of Celos on my redrawn map of The Lord of the Rings, see VII:322 n. 9."

Gilrain

A significant portion of the remarks on this river name was given in UT:242-45; but the discussion begins with a passage omitted from Unfinished Tales:

This resembles the name of Aragorn's mother. Gilraen; but unless it is misspelt must have had a different meaning. (Originally the difference between correct Sindarin ae and ai was neglected, ai more usual in English being used for both in the general narrative. So Dairon, now corrected, for Daeron a derivative of S. daer 'large, great': C.E. *daira < base DAY; not found in Quenya. So Hithaiglir on map for Hithaeglir and Aiglos [for Aeglos].){22} The element gil- in both is no doubt S. gil ‘spark, twinkle of light, star’, often used of the stars of heaven in place of the older and more elevated el-, elen- stem. (Similarly tinwe ‘spark’ was also used in Quenya). The meaning of Gilraen as a woman's name is not in doubt. It meant ‘one adorned with a tressure set with small gems in its network', such as the tressure of Arwen described in L.R. I 239.{23} It may have been a second name given to her after she had come to womanhood, which as often happened in legends had replaced her true name, no longer recorded. More likely, it was her true name, since it had become a name given to women of her people, the remnants of the Númenóreans of the North Kingdom of unmingled blood. The women of the Eldar were accustomed to wear such treasures; but among other peoples they were used only by women of high rank among the "Rangers", descendants of Elros, as they claimed. Names such as Gilraen, and others of similar meaning, would thus be likely to become first names given to maid-children of the kindred of the "Lords of the Dúnedain". The element raen was the Sindarin form of Q. raina ‘netted, enlaced'.

Against this last sentence Tolkien provided an etymological note:

C.E. base RAY ‘net, knit, contrive network or lace'; also [deleted: ‘catch',] ‘involve in a network, enlace'. Cf. Q. raima ‘a net';{24} rea and raita 1) ‘make network or lace'; raita 2) catch in a net';{25} [deleted: also raiwe ‘lace';] carrea (< cas-raya) a tressure'.{26} S. raef or raew (blend of Q. raima and raiwe) ‘net'; raeda- 'catch in net'; cathrae ‘tressure'. The word was only applied to work with a single thread; weaving with cross-threads or withes was represented by the distinct base WIG,{27} often in strengthened form waig-. The stems REB/REM were not "craft words" but verbal bases meaning ‘entangle, snare, trap (as hunters or fishers) with lines or nets". Cf. Q. rembe ‘net' (for catching), S. rem(m); Q. rembina 'entangled', S. remmen; Q. remi- ‘snare', remba- ‘net, entrap', remma ‘a snare', etc. Cf. S. Rem-mir-ath (‘group of gems in a net'), Pleiades.{28}

Of this note Christopher Tolkien writes: "Compare The Lord of the Rings, Appendix E (i), p. 393.{29}—Tressure, a net for confining the hair, is a word of medieval English which my father had used in his translation of Sir Gawain and the Green Knight (stanza 69): ‘the clear jewels / that were twined in her tressure by twenties in clusters', where the original has the form tressour."

The entry then concludes:

In Gilrain the element -rain though similar was distinct in origin. Probably it was derived from base RAN "wander, stray, go on uncertain course", the equivalent of Q. ranya. This would not seem suitable to any of the rivers of Gondor ...

The portion given in Unfinished Tales begins here (p. 242). The final sentence of the first extract from the discussion of Gilrain in UT:243 omits the ending; the whole sentence reads: "This legend [of Nimrodel] was well-known in the Dor-en-Ernil (Land of the Prince) and no doubt the name [Gilrain] was given in memory of it, or rendered in Elvish form from an older name of the same meaning" Also omitted was the paragraph following this sentence, which reads: "The flight of Nimrodel was dated by the chronologists at Third Age 1981. An error in Appendix B appears at this point. The correct entry read (still in 1963): "The Dwarves flee from Moria. Many of the Silvan Elves of Lorien flee south. Amroth and Nimrodel are lost.’ In subsequent editions or reprints ‘flee from Moria .. ' to ‘Silvan Elves has been for reasons unknown omitted." The correct reading of this entry has been restored in the latest edition (LR:1061). In addition, the first sentence of the following paragraph, introducing the passage with which the extract given in Unfinished Tales resumes (p. 243), reads: "At that time Amroth was, in the legend, named as King of Lorien. How this fits with the rule of Galadriel and Celeborn will be made clear in a precis of the history of Galadriel and Celeborn." Finally, the last sentence of the last paragraph given on UT:244 was omitted; it reads: "Communication was maintained constantly with Lorien."

A typescript note appended after the first sentence on UT:245, against the phrase "the sorrows of Lorien, which was left now without a ruler", and subsequently struck through by Tolkien, reads: "Amroth had never taken a wife. For long years he had loved Nimrodel, but had sought her love in vain. She was of Silvan race and did not love the Incomers, who (she said) brought wars and destroyed the peace of old. She would speak only the Silvan Tongue, even after it had fallen into disuse among most of the people. But when the terror came out of Moria she fled away distraught, and Amroth followed her. He found her near the eaves of Fangorn (which in those days drew much nearer to Lorien). She dared not enter that wood, for the trees (she said) menaced her, and some moved to bar her way. There they had long converse; and in the end they plighted their troth, for Amroth vowed that for her sake he would leave his people even in their time of need and with her seek for a refuge of peace. ‘But there is no such". The deleted note ends here, in mid-sentence. As Christopher Tolkien notes (UT:242), this passage is the germ of the version of the legend of Amroth and Nimrodel given in UT:240-42.

The discussion of Gilrain concludes (following the first paragraph given on UT:245) with this note:

The river Gilrain if related to the legend of Nimrodel must contain an element derived from C.E. RAN ‘wander, stray, meander'. Cf. Q. ranya 'erratic wandering', S. rein, rain. Cf. S. randír ‘wanderer' in Mithrandir, Q Rána name of the spirit (Máya) that was said to abide in the Moon as its guardian.

Uncertain, but probably from KIR ‘cut'. It rose in Lamedon and flowed westward for some way in a deep rocky channel.

For the element -ló see discussion of Gwathló above. But there is no record of any swamps or marsh in its course. It was a swift (and cold) river, as the element ring- implies.{30} It drew its first waters from a high snowfield that fed an icy tarn in the mountains. If this at seasons of snowmelting spread into a shallow lake, it would account for the name, another of the many that refer to a river's source.

Cf. the entry Ringló in the index to Unfinished Tales. This explanation of the name Ringló only arose in the course of the writing of this essay; for in the discussion of Gwathló that Tolkien struck out he had originally added this note:

In the revised discussion of Gwathlo (UT:263) this note was replaced by the following:

Tolkien then struck out the latter part of this note (from "since it formed a fenland" through the end), replacing it with a direction to see the final explanation of Ringló given above, in which the element lo is not derived from fenlands near the coast ("there is no record of any swamps or marsh in its course") but from the lake that formed at the river's source "at seasons of snowmelting" in the mountains.

Similarly the Morthond ‘Black-root', which rose in a dark valley in the mountains due south of Edoras, called Mornan,{31} not only because of the shadow of the two high mountains between which it lay, but because through it passed the road from the Gate of the Dead Men, and living men did not go there.

There were no other rivers in this region, "further Gondor", until one came to the Levnui, the longest and widest of the Five. This was held to be the boundary of Gondor in this direction; for beyond it lay the promontory of Angast and the wilderness of ‘Old Púkel-land' (Drúwaith Iaur) which the Númenóreans had never attempted to occupy with permanent settlements, though they maintained a Coast-guard force and beacons at the end of Cape Angast.

Levnui is said to mean ‘fifth' (after Erui, Sirith, Semi, Morthond), but its form offers difficulties. (It is spelt Lefnui on the Map; and that is preferable. For though in the Appendices f is said to have the sound of English f except when standing at the end of a word,{32} voiceless f does not in fact occur medially before consonants (in uncompounded words or names) in Sindarin; while v is avoided before consonants in English).{33} The difficulty is only apparent.

Tolkien then immediately embarks on a lengthy and elaborate discussion of the Eldarin numerals, which has been removed to an appendix below.

Following this discussion, Tolkien (continuing westward on the map from Levnui) reintroduced the name Adorn, and repeated the substance of his earlier remarks: "This river, flowing from the West of Ered Nimrais into the River Isen, is fitted in style to Sindarin, but has no meaning in that language, and probably is derived from one of the languages spoken in this region before the occupation of Gondor by Númenóreans, which began long before the Downfall." He then continued:

Several other names in Gondor are apparently of similar origin. The element Bel- in Belfalas has no suitable meaning in Sindarin. Falas (Q. falasse) meant ‘shore'—especially one exposed to great waves and breakers (cf. Q. falma ‘a wave-crest, wave). It is possible that Eel had a similar sense in an alien tongue, and Bel-falas is an example of the type of place-name, not uncommon when a region is occupied by a new people, in which two elements of much the same topographical meaning are joined: the first being in the older and the second in the incoming language.{34} Probably because the first was taken by the Incomers as a particular name. However, in Gondor the shore-land from the mouth of Anduin to Dol Amroth was called Belfalas, but actually usually referred to as i·Falas ‘the surf-beach' (or sometimes as Then-falas ‘short beach',{35} in contrast to An-falas ‘long beach', between the mouths of Morthond and Levnui). But the great bay between Umbar and Angast (the Long Cape, beyond Levnui) was called the Bay of Belfalas (Côf Belfalas) or simply of Bel (Côf gwaeren Bêl ‘the windy Bay of Bêl').{36} So that it is more probable that Bêl was the name or part of the name of the region afterward usually called Dor-en-Ernil ‘land of the Prince': it was perhaps the most important part of Gondor before the Númenórean settlement.

Christopher Tolkien writes: "With ‘the windy Bay of Bêl' cf. the poem The Man in the Moon came down too soon in The Adventures of Tom Bombadil (1962), where the Man in the Moon fell ‘to a foaming bath in the windy Bay of Bel', identified as Belfalas in the preface to the book.—This passage was struck through, presumably at once, since the next paragraph begins again ‘Several other names in Gondor are apparently of similar origin. A page of rapid manuscript found with the typescript essay shows my father sketching an entirely different origin for the element Bel-. I have referred to this text and cited it in part in Unfinished Tales (p. 247), observing that it represents an altogether different conception of the establishment of the Elvish haven (Edhellond) north of Dol Amroth from that given in Of Dwarves and Men (XII:313 and 329 n. 67), where it is said that it owed its existence to ‘seafaring Sindar from the west havens of Beleriand who fled in three small ships when the power of Morgoth overwhelmed the Eldar and the Atani'. The manuscript page obviously belongs to the same very late period as the essay, as is seen both from the paper on which it is written and from the fact that the same page carries drafting for the Oath of Cirion in Quenya (UT:305)." This manuscript page is given below in full; two notes that Tolkien made to the text are collected together at its end:

Belfalas. This is a special case. Bel- is certainly an element derived from a pre-Númenórean name; but its source is known, and was in fact Sindarin. The regions of Gondor had a complex history in the remote past, so far as their population was concerned, and the Númenóreans evidently found many layers of mixed peoples, and numerous islands of isolated folk either clinging to old dwellings, or in mountain-refuges from invaders (Note 1). But there was one small (but important) element in Gondor of quite exceptional kind: an Eldarin settlement.{37} Little is known of its history until shortly before it disappeared; for the Eldarin Elves, whether Exiled Noldor or long-rooted Sindar, remained in Beleriand until its desolation in the Great War against Morgoth; and then if they did not take sail over Sea wandered westward [sic; read "eastward"] in Eriador. There, especially near the Hithaeglir (on either side), they found scattered settlements of the Nandor, Telerin Elves who had in the First Age never completed the journey to the shores of the Sea; but both sides recognized their kinship as Eldar. There appears, however, in the beginning of the Second Age, to have been a group of Sindar who went south. They were a remnant, it seems, of the people of Doriath, who harboured still their grudge against the Noldor and left the Grey Havens because these and all the ships there were commanded by Cirdan (a Noldo). Having learned the craft of shipbuilding (Note 2) they went in the course of years seeking a place for havens of their own. At last they settled at the mouth of the Morthond. There was already a primitive harbour there of fisher-folk; but these in fear of the Eldar fled into the mountains. The land between Morthond and Serni (the shoreward parts of Dor-en-Ernil)

Note 1. Though none of the regions of the Two Kingdoms were before (or after!) the Númenórean settlements densely populated as we should reckon it.

Note 2: All Elves were naturally skilled in making boats, but the craft that were to make a long voyage over Sea, perilous even to Elven-craft until Middle-earth was far behind, required more skill and knowledge.

The manuscript page ends here, mid-sentence, and without reaching an explanation of the element Bel-. Christopher Tolkien writes: "It was perhaps a purely experimental extension of the history, at once abandoned; but the assertion that Cirdan was a Noldo is very strange. This runs clean counter to the entire tradition concerning him—yet it is essential to the idea sketched in this passage. Possibly it was his realization of this that led my father to abandon it in mid-sentence."

The typescript resumes with a replacement of the rejected passage on Belfalas (and now avoiding discussion of that problematic name):

Several other names in Gondor are apparently of similar origin. Lamedon has no meaning in Sindarin (if it was Sindarin it would be referred to *lambeton-, *lambetân-, but C.E. lambe- ‘language’ can hardly be concerned). Arnach is not Sindarin. It may be connected with Arnen on the east side of Anduin. Arnach was applied to the valleys in the south of the mountains and their foothills between Celos and Erui. There were many rocky outcrops there, but hardly more than in the higher valleys of Gondor generally. Arnen was a rock outlier of the Ephel Dúath, round which the Anduin, south of Minas Tirith, made a wide bend.

Suggestions of the historians of Gondor that arn- is an element in some pre-Númenórean language meaning ‘rock’ is merely a guess.{38} More probable is the view of the author (unknown) of the fragmentarily preserved Ondonóre Nómesseron Minaþurie (‘Enquiry into the Place-names of Gondor’).{39} On internal evidence he lived as far back as the reign of Meneldil, son of Anárion—no events later than that reign are mentioned—when memories and records of the early days of the settlements now lost were still available, and the process of naming was still going on. He points out that Sindarin was not well-known to many of the settlers who gave the names, mariners, soldiers, and emigrants, though all aspired to have some knowledge of it. Gondor was certainly occupied from its beginning by the Faithful, men of the Elf-friend party and their followers; and these in revolt against the ‘Adunaic' Kings who forbade the use of the Elvish tongues gave all new names in the new realm in Sindarin, or adapted older names to the manner of Sindarin. They also renewed and encouraged the study of Quenya, in which important documents, titles, and formulas were composed. But mistakes were likely to be made.{40} Once a name had become current it was accepted by the rulers and organizers. He thinks therefore that Arnen originally was intended to mean ‘beside the water, sc. Anduin'; but ar- in this sense is Quenya, not Sindarin. Though since in the full name Emyn Arnen the Emyn is Sindarin plural of Amon ‘hill', Arnen cannot be a Sindarin adjective, since an adjective of such shape would have a Sindarin plural ernain, or ernin. The name must therefore have meant ‘the hills of Arnen. It is now forgotten, but it can be seen from old records that Arnen was the older name of the greater part of the region later called Ithilien. This was given to the narrow land between the Anduin and the Ephel Dúath, primarily to the part between Cair Andros and the southern end of the bend of Anduin, but vaguely extended north to the Nindalf and south towards the Poros. For when Elendil took as his dwelling the North Kingdom, owing to his friendship with the Eldar, and committed the South Kingdom to his sons, they divided it so, as is said in ancient annals: "Isildur took as his own land all the region of Arnen; but Anárion took the land from Erui to Mount Mindolluin and thence westward to the North Wood", (later in Rohan called the Firien Wood), "but Gondor south of Ered Nimrais they held in common."

Arnach, if the above explanation is accepted, is not then related to Arnen. Its origin and source are in that case now lost. It was generally called in Gondor Lossarnach. Loss is Sindarin for ‘snow’, especially fallen and long-lying snow. For what reason this was prefixed to Arnach is unclear. Its upper valleys were renowned for their flowers, and below them there were great orchards, from which at the time of the War of the Ring much of the fruit needed in Minas Tirith still came. Though no mention of this is found in any chronicles—as is often the case with matters of common knowledge—it seems probable that the reference was in fact to the fruit blossom. Expeditions to Lossarnach to see the flowers and trees were frequently made by the people of Minas Tirith. (See index Lossarnach adding III 36,140;{41} Imloth Melui "sweet flower-valley", a place in Arnach). This use of ‘snow' would be specially likely in Sindarin, in which the words for fallen snow and flower were much alike, though different in origin: loss and loth, [the latter] meaning ‘inflorescence, a head of small flowers'. Loth is actually most often used collectively in Sindarin, equivalent to goloth; and a single flower denoted by elloth (er-loth) or lotheg.{42}

With Imloth Melui ‘sweet flower-valley cf Ioreth's mention of "the roses of Imloth Melui", LR:848. Against the Sindarin words loss and loth Tolkien made the following note:

S. loss is a derivative of (G)LOS ‘white'; but loth is from LOT. Sindarin used loss as a noun, but the strengthened form gloss as an adjective ‘(dazzling) white', loth was the only derivative of LOT that it retained, probably because other forms of the stem assumed a phonetic shape that seemed inappropriate, or were confusible with other stems (such as LUT ‘float'): e.g. *lod, *lûd. loth is from a diminutive lotse and probably also from derivative lotta-. Cf. Q. losse ‘snow', lossea ‘snow-white'; and late ‘a flower' (mostly applied to larger single flowers); olóte ‘bloom, the flowers collectively of a single plant'; lilótea ‘having many flowers'; lotse ‘a small single flower'; losta ‘to bloom', (t-t in inflexion > st.) Both Quenya and Sindarin retain for ‘snow' only the strengthened loss- since medial s between vowels suffered changes that made them unsuitable or clashed with other stems.{43}