In adapting the "Noldorin" material from the Etymologies to later Sindarin, many words must be subtly altered to conform with the revisions Tolkien did after writing Etym in the late thirties. One such change we described in the article Ae or Oe?, namely whether the primitive diphthong ai comes out as oe or ae in the Celtic-sounding language in Tolkien's mythos: In "Noldorin", the main rule was that ai became oe, as is reflected in most of the Etymologies. However, Tolkien changed his mind, and about the point of his turning Noldorin into Sindarin, he also decided that ai actually becomes ae in this language.
Another change, affecting a great number of words, involves the sounds that in Tolkien's Roman spelling of Noldorin/Sindarin are normally expressed by the digraphs lh and rh. In Quenya, the same sounds are normally spelt hl and hr. This is unvoiced L and R, sc. (dental) L and (trilled) R pronounced without vibration in the vocal cords. What Tolkien had in mind was evidently what may be described as "whispered" variants of normal Elvish L and R. In English, unvoiced l occurs only as a variant of normal l. (For instance, the l of please will normally be unvoiced; contrast the voiced l of lease - to borrow an example from An Introduction to Elvish.) But in Sindarin and Quenya, these unvoiced sounds are phonemes in their own right.
In the Noldorin of the Etymologies, rh- and lh- are common initial sounds. Primitive words and stems in r- and l- regularly yield Noldorin words in rh- and lh-, whereas the corresponding Quenya words have normal r- and l-, sc. the primitive sound unchanged. To quote just two examples, primitive rigê "crown" yields Noldorin rhî and Quenya rië (LR:383 s.v. RIG), whereas primitive lassê "leaf" produces Noldorin lhass and Quenya lassë (LR:367 s.v. LAS1). In Noldorin, one may say that the sounds lh and rh are simply how the phonemes /l/ and /r/ are realized when initial. However, the evidence is that Tolkien revised this. About the time Noldorin became Sindarin, he decided that in Sindarin as in Quenya, primitive initial r- and l- were simply unchanged. In the earliest LotR manuscripts, there are traces of the same system as in the Etymologies (for instance, in RS:438 we have Rhascarn as an early name of Caradhras, the Redhorn; this is clearly rhas(s) "horn" + carn "red"), but soon, the change becomes almost universal. Hence there are quite a few examples of Noldorin words in rh- and lh- corresponding to Sindarin words in r- and l-. We will look at some of them.
Noldorin rhom "horn, trumpet" (LR:384 s.v. ROM) reappears as rom in Sindarin (WJ:400); in both cases the primitive form was undoubtedly meant to begin in simple r- (probably *rombâ, cf. Quenya and Old Noldorin/Sindarin romba). Noldorin lham(b) "tongue" (LR:367 s.v. LAB) corresponds to Sindarin lam (WJ:394 cf. 416); both must be derived from primitive lambâ (or lambê when the meaning is "speech" rather than "physical tongue", but these coalesced in both Noldorin and Sindarin). Another example of the change from Noldorin lh- to Sindarin l- is seen in the word for "flower", Noldorin lhoth (LR:370 s.v. LOT(H)) corresponding to Sindarin loth (listed in the Silmarillion Appendix). Christopher Tolkien mentions a Noldorin eagle-name Lhandroval in LR:382 s.v. RAM; the new Sindarin form Landroval turns up in LotR3/VI ch. 4 ("[Gwaihir] lifted up Gandalf and sped away south, and with him went Landroval..."). Yet another example is provided by the Noldorin word lheithian "release, freeing" (LR:368 s.v. LEK); the published Silmarillion refers to "the Lay of Leithian, Release from Bondage" (chapter 19). The new form leithian also occurs on a slip accompanying the Etymologies itself; see below. As for further examples of the revision rh > r, we may compare the place-name Rhamdal (LR:390 s.v. TAL, TALAM) with the form Ramdal occurring in the Silmarillion. The Noldorin word for "cave" was rhond (or rhonn), listed in LR:384 s.v ROD. This clearly corresponds to later Sindarin rond, as in Hadhodrond as the Grey-elven name of Moria (WJ:414 indicates that Tolkien later decided that the stem was RONO rather than ROD, but this is immaterial). Compare also the Noldorin feminine name Rhian "Crown-gift" (LR:368 s.v. RIG) with Rían as the name of Huor's wife in the Silmarillion.
So there is no lack of evidence for the revision lh > l and rh > r. But don't the sounds rh- and lh- occur in later Sindarin as well? They do. But they are no longer to be derived from simple r-, l- in the primitive language, and they are no longer mere allophones of /r/ and /l/. In Appendix E to LotR, explaining his spelling conventions, Tolkien writes:
LH represents [L] when voiceless (usually derived from initial sl-). In (archaic) Quenya this is written hl, but was in the Third Age usually pronounced as l.
And similarly concerning rh:
RH represents a voiceless r (usually derived from older initial sr-). It was written hr in Quenya.
So now, Sindarin rh, lh (and Quenya hl, hr) are normally to be derived from sl-, sr- instead of simple l-, r- (for Tolkien had decided that they remained unchanged in both Sindarin and Quenya). The development sr > rh/hr is confirmed by a post-LotR source, Tolkien deriving Elvish words for "flesh" - Sindarin rhaw and Quenya hrávë - from primitive srâwê (MR:350). This new system differed from the linguistic scenario Tolkien had used in the Etymologies. When writing Etym back in the mid-thirties, Tolkien had primitive initial sr- and sl- yielding Noldorin thr- and thl-, whereas in Quenya initial sl- was merely simplified to normal l- (and sr- would probably likewise become simple r-, but Etym provides no example of this). This is evident from words like the verb thribi "to scratch" from the stem SRIP, or primitive slaiwâ "sickly" (stem SLIW) yielding Noldorin thlaew (or thloew). Initial thl eventually became fl, so thlaew later appeared as flaew (there are two other examples of this change: thliw "sickness", derived from the same stem SLIW, later becoming fliw, and thloss "whisper" from the stem SLUS later becoming floss). The Quenya cognates show simple l- (N thlaew/flaew = Q laiwa, N thiw/fliw = Q lívë, N thloss/floss = Q lussë). In the Quenya of the Etymologies, there are no words in hr- and hl-; the presence of these sounds in High-elven clearly had not entered Tolkien's conception at this time.
So, to summarize:
Old system (used in most of Etym):
Primitive
initial l- yields Noldorin lh- and Quenya l-
Primitive initial r- yields Noldorin rh- and Quenya r-
Primitive initial sl- yields Noldorin thl- (later becoming fl-)
and Quenya l-
Primitive initial sr- yields Noldorin thr- and Quenya *r-
(no examples)
But then, as he was reaching the LotR-stage, Tolkien decided that primitive l-, sl- and r-, sr- developed in the same way in both Quenya and "Noldorin" > Sindarin:
New system (used in LotR-style Quenya and Sindarin):
Primitive
initial l- remains l- in both Sindarin and Quenya
Primitive initial r- remains r- in both Sindarin and Quenya
Primitive initial sl- yields unvoiced L, spelt lh- in
Sindarin and hl- in Quenya
Primitive initial sr- yields unvoiced R, spelt rh- in
Sindarin and hr- in Quenya
The conclusion must therefore be this: In updating the "Noldorin" material from the Etymologies to LotR-style Sindarin, we shall have to change the initial consonant of words in lh- and rh- to normal l-, r-, while Noldorin words in thl- (fl-) and thr- must be altered to begin in lh- and rh- in Sindarin. We may also consider introducing hl-, not simple l-, in Quenya words derived from stems in SL- (and hr- in Quenya words derived from stems in SR- if Etym had provided any example of this, but the only relevant stem is SRIP, and no Quenya word is there listed).
As far as the "Noldorin"/Sindarin forms are concerned, there are traces (or should we say foreshadowings?) of the revised system in the Etymologies as well. In a number of cases, we have Noldorin words showing initial l- and r- instead of lh- and rh-, as if the New System is already lurking in the background: These words from the Etymologies can usually be adopted into Sindarin more or less as they are. The shifts between lh/l and rh/r seem unsystematic and probably simply reflect Tolkien's indecision. Of particular interest is the entry LEK (LR:368), where Tolkien first gave the word for "release, freeing" as lheithian. But on what Christopher Tolkien describes as "a slip accompanying these etymologies", this word becomes leithian instead: "Leithian release; cf. Lay of Leithian." It would seem that after writing the original entry, but before writing the accompanying slip, Tolkien made the revision. Other cases of l- instead of lh- already in the Etymologies are lalf and lalven (pl. lelf, lelvin) "elm-tree" (entry ÁLAM; contrast the form lhalwen in the entry LÁLAM, but even there, the pl. is given as lelwin instead of *lhelwin!), leithia "to release" (LEK, from the same accompanying slip that has leithian), lith "sand" (LIT), lond > lonn "path" (AK - but lhonn in the entry LOD!), Lothlann (place-name) (LUS - but Lhothlann in the entry LAD!). Cases of r- instead of rh- are râd "path, track", rado "to make a way, find a way", rant "lode, vein", rath "course, riverbed" (all listed under the stem RAT), Rathloriel (name of a river) (RAT, LÁWAR/GLÁWAR), Regornion "Hollin" (ERÉK), risto "rend, rip" (RIS), roch "horse" (ROK), rosto "to hollow out, excavate" (ROD).
Conversely, the Etymologies also lists a few Noldorin words in lh- and rh- that retain these initial sounds also in later Sindarin. We have, for instance, the word lhaw "(pair of) ears", in the Etymologies derived from an old dual lasû, a derivative of the stem LAS2 "listen". Normally, we wouldn't hesitate to emend lhaw to law in Sindarin. However, the form lhaw occurs in LotR itself, as part of the name Amon Lhaw "Hill of Hearing" or literally *"Hill of Ears" (mentioned near the end of the chapter The Great River in Volume 1). What are we to make of this? It seems that Tolkien in a few cases carried older Noldorin forms in lh-, rh- over into Sindarin, presumably simply because he liked the sound of them. Of course, this necessitated that he invented new etymologies for these words, referring them to primitive forms in sl- and sr- where l- and r- had sufficed in the Old System. It is clear that Tolkien in the post-LotR period derived lhaw from a stem SLAS, not LAS as in the Etymologies. This is evident from the related Quenya verb hlar- "listen", occurring in the late version of the Markirya poem (MC:222: Man hlaruva rávëa súrë[?] "Who shall hear the wind roaring[?]") In Quenya, initial hl- derives from earlier sl-. If we were to be wholly consistent, we would have to change l- to hl- in the Quenya words derived from LAS2 in the Etymologies (namely lár "pair of ears", the verb lasta- "listen" and the noun lasta "listening", plus the name Lastalaica "sharp-ears"). When the Noldorin word lhaw "(pair of) ears" kept its lh- in Sindarin, we must assume that the related singular lhewig "ear" would also be unchanged.
Another case of "Noldorin" lh- persisting in Sindarin is provided by the word lhûg "dragon, snake, serpent", first listed in the Etymologies (s.v. LOK, LR:370) but also reappearing in the Silmarillion Appendix (entry lok-), and there explicitly said to be Sindarin. But the latter source also states that the corresponding Quenya word was (h)lókë, whereas the Etymologies simply had lókë. A Quenya word hlókë requires a primitive form in sl- (in all likelihood *slôkê), so Tolkien has once again introduced a new etymology to keep a Sindarin word in its Noldorin form; for some reason he did not want to emend lhûg to lûg.
There are also a few cases of rh- persisting in Sindarin where we would normally expect r-. In the Etymologies, the entry RUSKÂ lists a Noldorin word rhosc "brown"; we would normally emend it to rosc in Sindarin, but in LotR we find Rhosgobel as the name of the home of one of the Wizards (Gandalf during the Council of Elrond referring to "Radagast the Brown, who at one time dwelt at Rhosgobel"). It can hardly mean anything else than "Brown Village", rhosc + gobel (for the latter element, see the entry PEL(ES) in the Etymologies). It seems that we must accept rhosc as the Sindarin word for "brown", and assume that it is derived from *SRUSKÂ rather than RUSKÂ.
Another word in rh- that Tolkien evidently could not part with is rhûn "east". It is listed in this form both in the Etymologies (LR:384 s.v. RÔ) and in LotR Appendix E. The related Quenya word rómen is also mentioned in both sources, which is somewhat curious: If Sindarin has rhûn from primitive *srônâ, and the same element *srô- is the origin of the first part of Quenya rómen as well, the Quenya word should have been *hrómen! The first part of rómen must be descended from *rô-, but we can't refer Sindarin rhûn to *rônâ, for this would produce Grey-elven *rûn instead. So despite the statement made by Christopher Tolkien in the entry rómen in the Silmarillion Appendix, that the Sindarin words rhûn, amrûn "were of the same origin" as Quenya rómen, there must be some stem-variation involved. The Quenya word must come from a stem RÔ-, while the Sindarin forms evidently reflect an S-prefixed variant of this stem: SRÔ-. Somewhere in the unpublished material, there is no doubt a note explaining how Sindarin can have rhûn when Quenya has rómen!
Perhaps the word rhass "precipice" should also be left unchanged. The stem is KHARÁS, so the Old Sindarin form would be *khrasse; the initial combination khr might very well come out as unvoiced rh also in Sindarin. (In Appendix E to LotR, Tolkien stated that Sindarin rh was "usually derived from older initial sr-", a wording indicating that there were also other primitive combinations that could yield this sound, though they were less common.) Moreover, emending rhass to rass would make it clash with the word for "horn".
We will try, then, to summarize how we must treat the Etymologies material to update it according to Tolkien's later discoveries about the evolution of Elvish.
1. In these Noldorin words, lh- should be changed to normal l- in Sindarin: lhaden (pl. lhedin) "open, cleared" (LAT), lhaeg "keen, sharp, acute" (LAIK), lhaes "babe" (LAP), lhagr "swift, rapid" (LAK2 - in LotR-style Sindarin, the form should be lagor rather than lagr), lhain < lhein "free, freed" (LEK - in LotR-style Sindarin, the form should be lain rather than lein), lhalorn "elm-tree" (LÁLAM), lhalwen "elm-tree" (LÁLAM; the pl. lelwin, in Noldorin surely an error for lhelwin, is already updated to Sindarin!), lhamb, lham "tongue" (LAB - the Sindarin form should be lam, attested in WJ:394 cf. 416, rather than lamb), Lhamthanc "forked tongue", serpent-name (STAK), lhanc "throat" (LANK), lhand "open space, level" (LAT), lhand, lhann "wide" (LAD - the Sindarin form should be land rather than lann), Lhandroval (eagle-name) (RAM; new form Landroval attested in LotR3/VI ch. 4), lhang "cutlass, sword" (LAG), lhant "clearing in forest" (LAT), lhasbelin "autumn" (KWEL, LAS1), Lhasgalen "Greenleaf", a name of Laurelin (LAS1; cf. Eryn Lasgalen "Wood of Greenleaves" as a new name of Mirkwood in LotR Appendix B), lhass "leaf" (LAS1), lhath "thong of [?leather]" (LATH), Lhathleg "sharp-ears" (name) (LAS2), lhathrado and lhathro "listen in, eavesdrop" (LAS2), lhathron "hearer, listener, eavesdropper" (LAS2), lhaug "warm" (LAW), lhaws "hair" (same gloss as Quenya lokse, not Noldorin lhoch!) (LOKH), lhebed "finger" (LEP/LEPET), lheben "five" (LEP/LEPEN/LEPEK), lhefi "to lick" (LAB; in Sindarin it should be spelt levi rather than lefi; the Noldorin 3. sg. present tense lhâf also mentioned in Etym would however simply become lâf), lhegin "swift, rapid" (LAK2), lheithian "release, freeing" (LEK; the new form leithian is found on a slip accompanying the Etymologies and also attested in the published Silmarillion), lheitho "to release" (LEK; the form leithia- found on the above-mentioned slip is perhaps to be preferred in Sindarin, rather than just changing lheitho to leitho, leitha-), lhend "tuneful, sweet" (LIND), lhevon or lhifnir "Elf remaining behind" (LEB/LEM) = Quenya Lembi (but the Noldorin words are sg. not pl.; the Lembi Tolkien later called Avari, Sindarin Evair, so these Noldorin words should perhaps be ignored rather than updated), lhimb, lhim "fish" (LIW - in LotR-style Sindarin, the form should be lim rather than limb), lhimlug "fish-dragon, sea-serpent" (LOK), lhimmid (pa.t. lhimmint) "moisten" (LINKWI; in Sindarin read perhaps limmida- rather than just limmid), lhimp "wet" (LINKWI), lhîn "pool" (LIN1), Lhîn Uial (place-name) (MUY), lhind, lhinn "air, tune" (LIN2 - in LotR-style Sindarin the form should be lind rather than linn, but -linn may be preferred in compounds like aerlinn), lhîr "row, range" (LIR2), lhoch "ringlet" (LOKH), lhoda "float" (LUT), lhoeb "fresh" (LÁYAK - the Sindarin form should be laeb rather than loeb, because of another revision Tolkien did), Lhoebelidh and Lhoebenidh "Green-elves" (LÁYAK - here we would also have to change oe to ae if the words were to be adapted to Sindarin, but for various reasons these "Noldorin" words are best ignored, and as Sindarin words for "Green-elves" they are in any case evidently obsoleted by Laegil, Laegrim, Laegel(d)rim, sg. Laegel, in WJ:385), lhong "heavy" (LUG1), lhonn "narrow path, strait" (LOD - the older Noldorin form, not given, would be lhond, cf. Quenya londë, and in LotR-style Sindarin we would also expect lond rather than lonn; the form lond is actually mentioned both in the entry LOD and in the entry AK), lhost "empty" (LUS), lhoth "flower" (LOT(H)), Lhothland > Lhothlann "empty and wide", name of a region (LAD; the "Sindarin" form Lothlann is attested in the entry LUS), lhû "a time, occasion" (LU), Lhuien = Quenya Lórien, name of a Vala (LOS), lhum "shade" (LUM), lhumren "shady" (LUM), lhûn "blue" (LUG2; however, the Silmarillion seems to indicate that Tolkien settled on luin as the Sindarin word for "blue"; a place-name Lhûn does occur in the published Silmarillion, whatever its derivation and meaning might be), Lhúndirien "Blue Towers", other name of Eredluin = "Blue Mountains" (LUG2; Christopher Tolkien cites an occurrence of Luindirien in another place, and this may be better Sindarin), Lhúnorodrim "Blue Mountains" (LUG2), lhunt "boat" (LUT), lhûth "spell, charm" (LUG2), lhútha- "to enchant" (LUG2), Lhúthien "enchantress" (fem. name = Lúthien) (LUG2).
2. The few lh-words from the Etymologies that should NOT be altered in Sindarin (because Tolkien evidently revised their derivation, referring them to stems in SL- instead of L-): lhewig "ear", lhaw "ears" (one person's pair of ears) (LAS2), lhûg "snake, serpent" (LOK; see lóke in the Silmarillion appendix for gloss). After the revision, we must assume that Tolkien introduced stems like SLAS- and SLOK- instead. [Update, 2007: In Parma Eldalamberon #17, pp. 77, 185, the stem SLAS yielding Sindarin lhaw is directly attested, and on p. 160, Sindarin lhûg is derived from (s)lôkô.]
3. The Noldorin words in thl- or fl-, derived from sl-, that in Sindarin should have lh- instead: thlaew, thloew "sickly, sick, ill" (SLIW; the Sindarin form should be lhaew rather than lhoew), thlê "fine thread, spider filament" (SLIG), thlein "lean, thin, meagre", pl. thlîn (SLIN; in LotR-style Sindarin, the sg. should be lhain rather than lhein), thlind, thlinn "fine, slender" (SLIN; the Sindarin form should be lhind rather than lhinn), thling "spider, spider's web, cobweb" (SLIG), thlingril "spider" (SLIG), thlinn, thlind "fine, slender" (SLIN), thliw, fliw "sickness" (SLIW; the Sindarin form should probably have a long vowel, lhîw), thloss, floss "a whisper or rustling sound" (SLUS/SRUS). Quenya words derived from the same stems should strictly have their initial consonant emended from l- to hl- if we were to carry Tolkien's revision all the way through: línë "cobweb", lia "fine thread, spider filament", liantë "spider" (all from SLIG), lívë "sickness", laiwa "sickly, sick, ill" (all from SLIW). The primitive word slindi "fine, delicate" should regularly have yielded Quenya lindë (old system) or hlindë (new system), but in LR:386 s.v. SLIN, Tolkien stated that it was blended with a distinct primitive word lindâ "sweet-sounding", that would produce Quenya linda (in both systems!) The meanings of lindâ and slindi merged to produce a general word for "fair". Hence slindi yielded neither lindë nor hlindë, and we needn't insist that linda should be emended to hlinda, since by its form this word is descended primarily from lindâ (simple primitive l- being unchanged in Quenya).
4. In these Noldorin words, rh- should be changed to normal r- in Sindarin: rhaen "crooked" (RAG), rhaes (probably a misreading for *rhass) "horn" (especially on living animal, but also applied to mountains; the Sindarin form rass is mentioned in LotR Appendix E) (RAS), rhaew "fathom" (RAK), rhafn "wing (horn), extended point at side, etc." (RAM), rhain < rhein "border" (REG; the Sindarin form should be rain rather than rein), rhamb, rham "wall" (RAMBÂ; the Sindarin form should be ram [listed in the Silmarillion Appendix] rather than ramb), Rhamdal "Wall's-end", name of a place in East Beleriand (TALAM; new form Ramdal attested in the published Silmarillion), Rhân "Moon" (RAN, THIL), rhanc pl. rhenc (archaic pl. rhengy) "arm" (RAK; the "non-archaic" pl. form rhenc should perhaps be rainc rather than renc in LotR-style Sindarin), rhandir "wanderer, pilgrim" (RAN), rhasg "horn" (especially on living animal, but also applied to mountains) (RAS), rhaud "metal" (RAUTÂ; meaning changed by Tolkien from "copper"), rhaudh "hollow, cavernous" (ROD), rhaug "demon" (RUK), rhaun "errant" (RAN), rhaw (1) "bank" (especially of a river) (RAMBÂ), (2) "wilderness" (RAB) (3) "lion" (RAW; the plural form rhui "lions" there listed would in Sindarin be roe rather than rui), rhedhi "to sow" (RED), rhein (< rhoein) "slot, spoor, track, footprint" (RUN; the Sindarin form should be rain rather than rein or roein), rhemb, rhem "frequent, numerous" (RIM; the Sindarin form should be rem rather than remb), rhenio (1) "to stray" (RAN), (2) rhenio "fly, sail; wander" (RAM; this latter "rhenio" is evidently a misreading for rhevio in Tolkien's manuscript, so the Sindarin form should be revio), rhess "a ravine" (RIS), rhest "cut" (noun) (RIS), rhî "crown" (RIG), Rhian ("crown-gift", fem. name) (RIG; the Sindarin form Rían occurs in the published Silmarillion), rhib- "to flow like a [?torrent]" (RIP; reading of gloss uncertain; the inflected forms rhimp and rhimmo should in any case also have r- rather than rh- in Sindarin), Rhibdath (and Rhimdath) "Rushdown" (river-name) (RIP), rhien, rhîn "crowned lady" = "queen", also adj. "crowned" (TÂ/TA3, RIG; the Sindarin form should be rîn rather than rien), rhîf "edge, hem, border" (RÎ; the Sindarin form should probably be spelt rîw rather than rîf), rhimb, rhim (1) "crowd, host" (RIM; the Sindarin form should be rim rather than rimb), (2) "pool or lake (in mountains)" (RINGI; again, the Sindarin form should be rim rather than rimb), rhingorn, rhin-gorn "circle" (RIN, KOR), rhinc "twitch, jerk, trick, sudden move" (RIK(H)), rhind, rhinn "circle", also adj. "circular" (RIN; the Sindarin form should be rind rather than rinn), rhing "cold" (RINGI); rhis "a ravine" (RIS; in Sindarin, the spelling riss rather than ris is perhaps preferable), rhîs "queen" (RIG), rhisto "cut" (vb) (RIS), rhitho "jerk, twitch, snatch" (RIK(H)), rhoeg "crooked, bent, wrong" (RÁYAK; the Sindarin form should be raeg rather than roeg), rhofal pl. rhofel "pinion, great wing (of eagle)" (RAM; in Sindarin, the sg. should be spelt roval rather than rofal, cf. the eagle-name Landroval, and the pl. should probably be rovail rather than rofel or rovel), rhom "horn, trumpet" (ROM; the new form rom is attested in WJ:400), rhomru "sound of horns" (ROM), rhond, rhonn "cave" (ROD; the Sindarin form should be rond rather than ronn; this form is attested in WJ:414, thought Tolkien had introduced a new etymology), rhoss (1) "rain" (ROS1), (2) "polished metal" (RUS), rhû "loud sound, trumpet-sound" (ROM), rhuiw or rhui "hunt, hunting" (ROY1), rhŷn " 'chaser', hound of chase" (ROY1).
5. The few rh-words from the Etymologies that should NOT be altered in Sindarin (because Tolkien evidently revised their derivation, referring them to stems in SR- instead of R-): rhosc "brown" (RUSKÂ), rhûn and rhufen "east" (RÔ; rhufen is listed as an archaic or poetic word; it is better spelt rhuven according to the system Tolkien used in LotR). Perhaps rhass "precipice" (KHARÁS) should also be left unchanged; see above.
6. The Noldorin words in thr-, derived from sr-, that in Sindarin should have rh- instead: thribi "to scratch" (SRIP) and thross "a whisper or rustling sound". (For the latter, see the entry SLUS, SRUS; in this case the stem is obviously SRUS - whereas the Quenya words lussë and lussa-, "whisper" as noun and verb respectively, must be derived from SLUS: Hence they should strictly become hlussë and hlussa- if we were to carry Tolkien's revision all the way through.)