A noun for “branch” in The Etymologies of the 1930s derived from ᴹ✶golbā under the root ᴹ√GOLOB (Ety/GÓLOB). It was an element in the name N. Gurutholf “Wand of Death” (Ety/ÑGUR).
Conceptual Development: The Gnomish Lexicon of the 1910s had G. olf or olfin(g) “branch, wand, stick” (GL/62), an element in the earlier name G. Gurtholfin “Wand of Death” (GL/41; LT2/83). The Gnomish Lexicon originally had two words: olf “branch” and olfin “wand”, but these were deleted and merged into one word with two variants (GL/41). See ᴱN. alt for other “branch” words from the 1910s and 20s.
In Tolkien’s later writings, N. Gur(u)tholf became S. Gurthang “Iron of Death” (S/210; WJ/83). However, there was a word Q. olba “branch” in notes from 1968, derived from √OLOB (PM/341).
Neo-Sindarin: The 1968 Quenya word might mean Noldorin golf should be updated to [ᴺS.] ✱olf, but I prefer to assume there was a variant root √(G)OLOB which would allow us to retain golf “branch” for purposes of Neo-Sindarin.
An impersonal verb appearing as N. eil “it is raining” in The Etymologies of the 1930s derived from primitive ᴹ✶ulyā- (> œil > eil) under the root ᴹ√ULU “pour, flow” (Ety/ULU; EtyAC/ULU). This verb is abnormal in that its final a disappeared rather than surviving as it usually did for derived verbs, giving eil rather than ✱elia-, ✱eilia- or ✱eila-. However, as an impersonal verb there would never be any pronominal suffix to help preserve the final a by analogy, which probably explains the vowel loss.
Conceptual Development: The verb for “rain” in the Gnomish Lexicon of the 1910s was G. ubra-, probably related to G. ub “wet, moist, damp” (GL/74).
Neo-Sindarin: In Sindarin, the likely developments of primitive ✶ulyā would be to ✱oly(a) > ᴺS. uil “it rains”; I believe this form was first suggested by Helge Fauskanger in his Parviphith Edhellen wordlist. Compare ᴺS. uil to: S. ruin “fiery red” < (perhaps) ✱runyā and S. fuir “north” < (perhaps) ✱phoryā, and see the entry on how [[s|[œi] became [ui] or [y]]] for further discussion. Any inflected forms would probably restore the stem, such as (hypothetical) intransitive past and future forms ✱eilias “it rained” and ✱eiliatha “it will rain”).