Greek question

[Peter &/or Graham]

Patrick said:>
>The entry in Pokorny that comes closest to being related to the Greek forms
>is *neuk(h)-, 'dark', from earlier *neugh- + -s/t- (metathesis in Latin

Mann suggests *nokw-t-s alternating with nokw-t-is,  and points in evidence
to Lith. nakvoti, and hesitantly also to Hittite nekuc.  He gives other
parallel examples of *-okw- > -ukw- to explain the Greek -u- vocalism.
Both the -i- stem and the consonant stem are found in Vedic.

He is working, as you see, with a PIE vowel system which includes /o/.

Peter