This is the Quenya word for “head”, with a stem form of car- because medial s generally became z and then r, but the s was preserved when final. This word can refer to the head of people and animals, as well as the metaphorical “head” (or top) of other things, in much the same way that Q. tál “foot” can refer to their base.
Conceptual Development: This word was established very early in Tolkien’s writing, being derived from the root ᴱ√KASA “head” all the way back in the Qenya Lexicon of the 1910s (QL/45), but its exact form varied as Tolkien changed his mind on the phonetic development of s in Quenya. Its form in the Qenya Lexicon was in fact ᴱQ. kar (kas-), since in Early Qenya period medial s survived and it was final s that became r (PE12/26). This kar (kas-) was the usual word for head in the 1910s and 20s, but in the typescript version of the Early Qenya Grammar Tolkien instead revised it to ᴱQ. kas (kast-) “head” (PE14/72 and note #5).
In noun declensions from the late 1920s and early 1930s, Tolkien instead had cas (car-), reflecting a conceptual shift in the phonologic development of s (PE13/112-113; PE21/22). However, for reasons unclear, the form ᴹQ. kár (kas-) was restored in The Etymologies written around 1937 under the root ᴹ√KAS “head” (Ety/KEM), despite s > z > r being the normal medial phonetic development in this period (PE19/33). This abnormal form slipped into The Lord of the Rings itself as part of the name Q. Eldacar “Elfhelm” (LotR/1038).
Tolkien generally used the form cas for “head” in his later writings (PE19/103; VT49/17), but in his notes on Words, Phrases and Passages from the Lord of the Rings from the late 1950s or early 1960s, Tolkien was forced to contrive another explanation for Eldacar:
> What is -kar in names. How could it stand for helm? E.g. as stem ✱kāsā (√KAS, head) would give kāra, but in compound forms -kāsă > -kas. Would not an ă be lost before voicing of s or at least before z > r (PE17/114).
In this note Tolkien considered having Q. carma “helm” < kas-mā, but discarded the idea since he felt karma “tool or weapon” < KAR “do, make” + mā was the more likely meaning. He then said “Eldă|kāzā in compounds to -kār(ă) > -kar” despite its phonological implausibility, and indeed kāza/kára appeared in a discussion of helms within 1964 notes on Dalath Dirnen (DD: PE17/188).
As for the sense “top”, there is better evidence for it among Tolkien’s earlier writings, such as the glosses “head, top” in Early Qenya Grammar of the 1920s (PE14/79) and the early-1930s allative form kasta “up (to the top)” (PE21/22). I see no reason to assume this alternate meaning did not survive in Tolkien’s later conception of the language.
cas ("k")"head" (VT49:17), cf. also deleted [cas] ("k")noun "top, summit" (VT45:19). This noun should evidently have the stem-form car-. See cár.