Comments:VEWT 220, EDT 587, 606, ЭСТЯ 6, 161-163. Modern reflexes point quite clearly to *-g- (less probably - *-b-), so the attested form with -δ- must be a phonetic aberration. Interference with OT qoduz 'femme s{eu}le' (suggested in Tekin 1969 and ЭСТЯ ibid.) is not excluded, but cf. also the notes below.
