Notes: A typical case of the structure HRVMCV̆ (in PN *n- < *m- before a front vowel). The Khin. form raises some problems: it must be a metathesis (ḳʷar < *raḳʷ; the supposed form *raḳʷ would be a regular development < *rVq̇_wV which must be a result of secondary labialisation < *rVmq̇_V having occurred early in the history of the Khinalug language).
The EC and WC forms were compared in Trombetti 1923, Ломтатидзе 1955, Mészarós 1934, Абдоков 1983. The comparison seems reasonable enough, but we must note that to justify it phonetically we must suppose a metathesis *Hrēmq̇_ĭ > *Hmērq̇_ĭ in early Proto-WC (otherwise *-m- would have been lost); it is even possible that the Adygh languages reflect the normal *Hrēmq̇_ĭ (no nasal is preserved in PAK), while Ub. and PAT reflect the metathesized variant. Labialisation in PWC is probably assimilatory (after a labial consonant); palatalisation is regular.