ᴱ√WĀ “blow; noise of wind, echoic representation of sound of wind”
This and similar roots were the basis for “wind” words for much of Tolkien’s life, especially in Sindarin and its precursors. Its first appearance was unglossed ᴱ√GWĀ in the Qenya Lexicon of the 1910s with derivatives like ᴱQ. ’wā “wind” and ᴱQ. ’wanwavoite “windy”, where presumably the ’ indicated the lost initial g; there were also two erased variants of the root ᴱ√WĀ and ᴱ√WA’A (WAƷA?) (QL/102). In the contemporaneous Gnomish Lexicon it had derivatives like G. gwâ “wind” and G. gwavwed “windy” (GL/43).
In The Etymologies of the 1930s Tolkien gave the root as ᴹ√WĀ “blow” with extended variants ᴹ√WAIWA and ᴹ√WAWA and derivatives like ᴹQ. vaiwa/N. gwaew “wind” (Ety/WĀ). ᴹQ. ván/N. gwaun < ᴹ√WĀ-N also seem to be related (Ety/WA-N; EtyAC/WA-N), perhaps based on the noise the goose makes. The root √WĀ appeared a number of times in Tolkien’s later writings with glosses like “blow” (PE17/33), “[used?] of noise of wind” (PE17/34), “echoic representation of sound of wind” (NM/237), and in the variant form √WAY “blow, or be disturbed” (PE17/33) or √WAY “blow (as of wind)” (PE17/154, 189). √WAY was its most common variant form, but it had many others such as √WAW, √WIW, √SWA, √SWAW and √SWAR.
√WĀ/AWA “away (from); go (away), depart, pass away, move (from speaker); before (of time), ago; [ᴹ√] forth, out”
This invertible root and ones like it were the basis for “away” words for much of Tolkien’s life. The earliest iteration was ᴱ√AVA “go away, depart, leave” in the Qenya Lexicon of the 1910s with derivatives like ᴱQ. au “away from” and ᴱQ. avanwa “going, passing, nearly gone” (QL/33). This early root remanifested as ᴹ√AB “go away, depart, leave” in The Etymologies of the 1930s, but the gloss of that root was revised to “refuse, deny” (Ety/AB). As a replacement, Tolkien introduced ᴹ√AWA “away, forth; out” with derivatives like ᴹQ. ava “outside”; Tolkien also considered deriving a privative prefix ᴹQ. ava- from this root (Ety/AWA).
The root √AWA was mentioned many times in Tolkien’s later writings, along with its inverted variant √WĀ, usually with the sense “away (from)” or a verbal sense “go (away), depart, pass away”. Its most detailed description appeard in the Quendi and Eldar essay of 1959-60, where Tolkien said:
The element *AWA … referred to movement away, viewed from the point of view of the thing, person, or place left. As a prefix it had probably already developed in CE the form *au-. The form *awa was originally an independent adverbial form, but appears to have been also used as a prefix (as an intensive form of *awa-, *au-). The form *wā- was probably originally used as a verbal stem, and possibly also in composition with verbal stems (WJ/361).
In this same document Tolkien said of Sindarin that:
The only normal derivative [of √AWA] is the preposition o, the usual word for “from, of”. None of the forms of the element *awa are found as a prefix in S, probably because they became like or the same as the products of *wō, *wo (WJ/366).
Indeed, most of the attested derivatives of this root are in Quenya, but there are a couple in Sindarin, such as the aforementioned S. o, as well S. gwanwen “departed” (WJ/378) and the verb S. gwae- “go”, probably only in the limited sense “depart” (PE17/148).
In late notes from 1969 Tolkien gave the root √AWA the sense “before or ago (of time)” (PE22/167 note #117; PE22/168), but I suspect this was a transient idea.
ᴹ√WAƷ “stain, soil”
A root in The Etymologies of the 1930s glossed “stain, soil” with derivatives like ᴹQ. vakse/N. gwass “stain” and ᴹQ. vára/N. gwaur “soiled, dirty” (Ety/WAƷ). Tolkien also gave it an alternate form ᴹ√VAG (EtyAC/WAƷ). The 1930s name N. Iarwath or Iarwaeth “Blood-stained” was based on this root (Ety/WAƷ, YAR; WJ/83). In the 1950s Tolkien change this name to S. Agarwaen (WJ/142), and the second element of this name (S. *gwaen) might still be based on ᴹ√WAƷ.
ᴹ√WAN “depart, go away, disappear, vanish”
A root in The Etymologies of the 1930s glossed “depart, go away, disappear, vanish” with derivatives like ᴹQ. vanwa “gone, departed, vanished, lost, past” and N. gwann “departed, dead” (Ety/WAN). It was revised in pencil to ᴹ√VAN. The 1930s root is probably a later iteration of unglossed ᴱ√VAHA in the Qenya Lexicon of the 1910s which was an elaboration of ᴱ√AVA “go away, depart, leave”, and had derivatives like ᴱQ. vand- “way, path”, ᴱQ. vandl “staff”, and ᴱQ. vanwa “gone, on the road, past, over, lost” (QL/99). It also had derivatives in the contemporaneous Gnomish Lexicon such as G. bâ “away, off”, G. bad- “travel”, and G. bang “staff” (GL/21).
In later writings, Q. vanwa was derived from the invertible root √WĀ/AWA; in the 1930s there is no indication that ᴹ√AWA “away” was invertible, so perhaps inversion √WĀ can be considered the conceptual descendent of ᴹ√WAN and ᴱ√VAHA (WJ/366). For a time in the 1940s and 1950s, Q. vanwa was derived instead from √BA(N) “go, proceed” (PE22/97; PE17/16), but this root was abandoned in 1959. Thus, perhaps the full evolution was 1910s ᴱ√VAHA >> 1930s ᴹ√WAN >> >> 1940s-50s √BA(N) >> 1960s √WĀ.
ᴹ√WAR “give way, yield, not endure, let down, betray”
A (Noldorin only?) root in The Etymologies of the 1930s glossed “give way, yield, not endure, let down, betray” with derivatives like N. awarth “abandonment”, N. gwarth “betrayer”, and N. gweria- “betray” (Ety/WAR). Tolkien said it was the opposite of ᴹ√BOR “endure” (EtyAC/WAR) which had various Noldorin derivatives having to do with loyalty (Ety/BOR).
ᴱ√WARA “care for, guard, watch (over)”
This root appeared in the Qenya Lexicon of the 1910s as ᴱ√QARA “care for, guard, watch (over)” along with derivatives ᴱQ. qāra “watch, ward”, ᴱQ. qārele “watchfulness, anxiety”, and Gnomish variant gwar- (QL/76). In the contemporaneous Gnomish Lexicon G. gwar- appear with the gloss “watch (all senses)” with an apparent primitive form u̯ar “guard” (GL/46). This may represent a conceptual shift from primitive {ᴱ✶gʷar- >>} ᴱ✶war-, with earlier Qenya kʷar- being the result of the usual change whereby initial voiced stops were unvoiced [gʷ- > kʷ-]. Further evidence of this shift may be found in ᴱQ. Varavilindo, Qenya cognate of G. Gwarbilin “Birdward(en)” (GL/70), where the Qenya form was the result of w becoming v.
The most notable derivative of this root was G. gwareth “watch, guard, ward” as in G. Amon Gwareth “Hill of Watch” (LT2/158), a name that retained this form and meaning into Silmarillion drafts of the early 1930s (SM/137). In the mid-1930s, Tolkien considered changing the name to N. Amon Thoros (LR/56), but ultimately retained this name as S. Amon Gwareth into the 1950s, though without translation (WJ/200; S/126). It is thus possible this early root ᴱ√(G)WARA “guard” survived, but it is more likely that the name Amon Gwareth survived without being given a proper etymology in the Elvish languages as Tolkien imagined them in the 1950s and 60s.
√WATH “[ᴹ√] shade”
The root ᴹ√WATH “shade” first appeared in The Etymologies of the 1930s with derivatives like N. gwath of the same meaning (Ety/WATH). Primitive ✶wath “shadow” was mentioned in notes on Words, Phrases and Passages from The Lord of the Rings from the late 1950s as the basis for S. gwath (PE17/41), and primitive root √WATH appeared in notes on The Rivers and Beacon-hills of Gondor from the late 1960s, again as the basis for S. gwath and similar words like S. gwathren “shadowy, dim” and Q. vasar “veil”, the latter from extended form waþar [√WATHAR] which was apparently the only survival of the shorter root in Quenya (VT42/9-10). This extended root might also be the basis for the (archaic?) region name Q. Avathar “Shadows” from The Silmarillion (S/73-4; MR/284).
ᴹ√WAY “enfold”
The earliest iteration of this root was ᴱ√VAẎA “enfold, wind about” from the Qenya Lexicon from the 1910s with derivatives like ᴱQ. vaile “covering”, ᴱQ. vaima “wrap, robe”, and ᴱQ. vaita- “wrap” (QL/100). The contemporaneous Gnomish Lexicon had derivatives like G. baidha- “clothe”, G. bail “sheath, case, cover; pod”, and G. bain “clad”, based on the primitive form Baʒ̔- (GL/21). This likely meant the medial consonant of the root was actually the voiced palatal spirant [ç], which Tolkien generally represented as ẏ in Qenya roots (PE12/15-16).
{ᴱN. bain >>} ᴱN. bai “clad” and ᴱN. bail “sheath” from Early Noldorin Word-lists of the 1920s were probably related (PE13/138). In The Etymologies from the 1930s the root appeared as ᴹ√WAY “enfold” with derivatives like ᴹQ. vaita-/N. gwaeda- “enfold” and N. ui² “envelope” (Ety/WAY; EtyAC/WAY). In this entry the form was revised to ᴹ√VAY (Ety/WAY). Tolkien also indicated that it was “confused in later Q with BAYA”, perhaps an attempt to salvage the bai- forms of Gnomish and Early Noldorin, but this sentence was struck out and there is no other sign of ᴹ√BAY in The Etymologies.
Neo-Eldarin: In The Etymologies, the Noldorin words for “clothing” seem to have moved to a new root, ᴹ√KHAP “enfold”, but this root had no Quenya derivatives. I think it is best to reconceive of the Early Qenya derivatives of ᴱ√VAẎA related to “clothing” as derivatives of ᴹ√WAY instead. The Gnomish derivatives of ᴱ√VAẎA can’t be easily salvaged, however, since the later roots changed to make it difficult produce their initial b.