This root and ones like it were the basis for time words throughout Tolkien’s life. The earliest appearance of this root was as ᴱ√LUHU or ᴱ√LU’U (the latter marked by Tolkien with a “?”) in the Qenya Lexicon of the 1910s, unglossed but with derivatives like ᴱQ. lú “24 hours, day” and ᴱQ. lúme “time” (QL/56). It also had derivatives in the contemporaneous Gnomish Lexicon like G. lu “occasion, time” and G. lust “time; tide; weather” (GL/55). It appeared as unglossed ᴹ√LU in The Etymologies of the 1930s with derivatives like ᴹQ. lú “a time, occasion”, ᴹQ. lúme “time”, and their Noldorin cognate N. lhû (Ety/LU).
The Etymologies had a distinct root ᴹ√ULU “pour, flow” (Ety/ULU). It was not explicity connected to ᴹ√LU in this 1930s document, but in notes from around 1959 Tolkien said lūmē “time” was derived from √ULU “flow” (PE17/168), so it is likely that √LU is an inversion of this more basic root.
A hypothetical root implied by the primitive word ✶lubbu “a clumsy piece or lump” appearing in both the Outline of Phonetic Development (OP1) from the 1930s and the Outline of Phonology (OP2) from the early 1950s serving to illustrate the unvoicing of double stops in Quenya: ✶lubbu > Q. luppo (PE19/45, 92). It may be related to earlier ᴱQ. ulumpe “camel”; although its root was given as ᴱ√ULUN(T) in the Qenya Lexicon of the 1910s, Tolkien gave ulumpe- in parenthesis beside the root, indicating it was probably an elaboration on unattested *√LUP (QL/97). It might also be connected to G. lub “fat, fat flesh” < ᴱ✶lūpe as well as G. lubi “corpulent” (GL/55), though that is more of a stretch. I think it is worth positing a variant root ᴺ√LUP “hump” for Neo-Eldarin in order to salvage Early Qenya camel words.
ᴹ√LUG¹ “heavy” “be heavy”
A root in The Etymologies of the 1930s glossed “be heavy”, with the derived adjectives ᴹQ. lunga and N. lhong of the same meaning (Ety/LUG¹). Given the appearance of G. lung “heavy; grave, serious” in the Gnomish Lexicon of the 1910s, I think the idea for this root dates back to this period, though the related verb G. luntha- “balance, weigh” indicates the Early-period root may have been *ᴱ√LUŊU instead (GL/55). Tolkien’s continued used of Q. lungu- and S. -lung for “heavy” in his later writings indicates its ongoing validity (S/185; PE17/162; VT47/19).
√LUK “haul, drag; [ᴹ√] magic, enchantment”
This root appeared a number of times in Tolkien’s writing over the years, but never with the same meaning. In the Qenya Lexicon of the 1910s, unglossed ᴱ√LUKU appeared with derivatives like ᴱQ. lúke “slime” and ᴱQ. lukso “mud” (QL/56). The Gnomish words G. ûg “mud” and G. ûgrin “muddy” in the contemporaneous Gnomish Lexicon may be related, but if so also indicate a shift in the root to *ᴱ√UKU (GL/74).
In The Etymologies of the 1930s the root ᴹ√LUK “magic, enchantment” appeared with derivatives like ᴹQ. lúke “enchantment” and N. lhûth, and the root served as the basis for the name Ilk. Lúthien, translated as “Enchantress” (Ety/LUK). However, in later writings the name S. Lúthien was given a new etymology as a feminization of S. lûth “blossom, inflorescense” (PE17/15).
The root √LUK itself reappeared in notes from around 1968 with the gloss “haul, drag”, serving as the basis for Q. lunka “heavy transport wain (wagon)” (PE17/28; VT43/19). This sense of the root might have an earlier origin, since the verb ᴱQ. luk- “pull” in Early Qenya word lists from the 1920s has a similar meaning (PE16/134).
Another set of possibly related forms are Q. lucassemmar, Q. luciemmar, or Q. luhtammar “*our debts, our trespasses” from Quenya prayers of the 1950s, as suggested by Wynne, Smith, and Hostetter (VT43/19). These appear to be derived from √LUK, but don’t seem to be connected to any of the attested meanings of this root. In any case, these words for “trespasses” were replaced in later versions of the prayers, becoming Q. rohtammar >> Q. úcaremmar (VT43/19).
For purposes of Neo-Eldarin, √LUK “haul, drag, *pull” is probably the most enduring meaning of this otherwise variable root, but the sense ᴹ√LUK “magic, enchantment” is also very popular, being the basis for some of the better known Elvish words for “magic”. I personally retain both senses, with “magic” by way of analogy for “pulling” on the physical world to affect change. There is a similar analogy to “breath” for the “emission of power (of will or desire)” by analogy with breath fogging a cold surface; see the entry on Q. súlë for discussion. However, this interpretation of √LUK is pure invention on my part to justify retaining the 1930s words for magic.
√LUM “shadow, darkness”
This root and ones like it were the basis for shadowy things throughout Tolkien’s life, but went through a number of minor conceptual shifts. The earliest appearance of this root was as ᴱ√LUVU in the Qenya Lexicon of the 1910s, with derivatives like ᴱQ. lumbo “dark lowering cloud” and ᴱQ. lūre “dark weather” (QL/57). In the Poetic and Mythological Words of Eldarissa written afterwards, Tolkien gave the root as ᴱ√LUB with a similar set of derivatives (PME/57); phonological developments in both Early Qenya and Gnomish make it very difficult to distinguish ancient voiced stops [b] from voiced spirants [β]. Sign of this root can also be seen in Gnomish words G. lum or glum “cloud”, G. lumbri “foul weather”, and G. luv- “hang, lower, of clouds” (GL/55).
The derivatives of this root in the 1910s seem to connect more specifically to dark weather, but in the The Etymologies of the 1930s the root reappeared as ᴹ√LUM with derivatives having to do mainly with shadow, such as ᴹQ. lumbe “gloom, shadow” and N. lhum “shade” (Ety/LUM). These in turn served as the basis for N. Hithlum and ᴹQ. †Hísilumbe >> ᴹQ. Hisilóme interpretted in this period as “Mist-and-Dusk” (LR/406). In earlier writing the second element of ᴱQ. Hisilóme “Misty-gloom” was derived from ᴱ√LOMO (QL/55), whereas in 1964 notes Hithlum was designated “North Sindarin” and given a new etymology as a direct loan from its Quenya equivalent, and thus no longer connected to √LUM (PE17/133).
The last appearance of the root √LUM itself in currently published material was as √LUM or √LUB “shadow, darkness” with derivatives Q. lumbo “dark, shade” and Q. lumbule “shadow” (PE17/168). Since √LUM has the fewest collisions with other roots (compare: ᴹ√LUB “weary”, √LUB¹ “bend” and *√LUB³ “lump”), I recommend using it for purposes of Neo-Eldarin. However, Q. Luvailin “Shadowmere” (RC/217) is almost certainly derived from √LUB.
A root in The Etymologies of the 1930s whose derivatives had to do with empty things, such as ᴹQ. lusta “void, empty”, N. lhoss “wilderness” and N. lhost “empty” (Ety/LUS). The last of these was an element in the name N. Lhothlann or Lothland “Empty and Wide” (Ety/LAD; LR/264), and Tolkien’s continued use of S. Lothlann in later versions of the Silmarillion indicate the ongoing validity of this root (S/123).
√LUT “float, [ᴹ√] swim”
This root was connected to floating and boats for all of Tolkien’s life. It first appeared as unglossed ᴱ√LUTU in the Qenya Lexicon of the 1910s with derivatives like ᴱQ. lunte “ship” and ᴱQ. lutu- or lutta- “flow, float” (QL/57). It also had derivatives in the contemporaneous Gnomish Lexicon such as G. laud “flood; high tide; tide, motion of the sea”, G. lud- “flow, stream, float”, and G. lunta “a ship” (GL/53, 55). It appeared as ᴹ√LUT “float, swim” in The Etymologies of the 1930s with derivatives ᴹQ. lunte/N. lhunt “boat” and N. lhoda- “to float” (Ety/LUT). The root √LUT “float” was mentioned in passing within notes from the late 1960s having to do with The Rivers and Beacon-hills of Gondor (VT42/18).
The Elvish words for “blue” remained very similar throughout Tolkien’s life, but underwent a number of minor conceptual shifts. The word ᴱQ. lūne “blue, deep blue” appeared in the Qenya Lexicon of the 1910s in a collection of words along with ᴱQ. lūle “blue stone, sapphire”, but no root was given (QL/55). The word for “blue” in the contemporaneous Gnomish Lexicon was G. luim (GL/55). In The Etymologies of the 1930s, the root for “blue” was ᴹ√LUG² with derivatives ᴹQ. lúne and N. lhûn (Ety/LUG²).
Meanwhile, the root ᴹ√LUY appeared in The Etymologies with derivatives ᴹQ. luina and Dor. luin “pale” (EtyAC/LUY), probably connected to ᴱN. Draugluin “Werewolf Pale” from the Lays of Beleriand of the 1920s (LB/205). But in The Etymologies the root ᴹ√LUY was rejected, and Dor. luin “pale” was reassigned to ᴹ√LUG² and then revised in form to Dor. lūn (Ety/LUG²; EtyAC/LUG²).
In addition, there was already evidence of a conceptual shift in the Noldorin words for blue in the 1930s, with the name N. Eredluin “Blue Mountains” being given as an alternative to N. Lhúnorodrim and N. Lhúndirien “Blue Towers” (Ety/LUG²), the latter appearing as N. Luindirien in contemporaneous Silmarillion narratives (LR/267). By the 1950s and 60s, the Sindarin and Quenya words for “blue” had firmly become S. luin (Let/448; S/54; UT/390) and Q. luinë (LotR/377; PE17/66, 71). The root √LUY “blue” appeared in notes from the late 1960s serving as the new basis for these “blue” words (VT48/23-24, 26).
All this made a mess for the river name S. Lhûn (LotR/1134) from The Lord of the Rings which was a remnant of Tolkien’s earlier ideas, and he struggled to find a new basis for that name as discussed by Patrick Wynne in his article on The Problem of Lhûn (VT48/26-29).