From: tgpedersen
Message: 49525
Date: 2007-08-15
>Actually Beekes does propose *tpolis ("[However, a metathesis tp > pt
> > Perhaps polis/ptolis < *tpolis, like pelea/ptelea "elm" <*tpeleya
> > (cf. Slavic topol-) and khamai (<*dHgHmai)/khthon (<*dHgHom)
> ("1a. pt may represent a single phoneme py, as we saw in B 1.
> Exx. (Fur. 315ff): gnup- / gnupt- (gnupet-); kolúmbaina / kolúbdaina;
> kíbalos / kíbde:s; lúpe: / lúpta; without variants note króssophthon,
> sarúphthei~n. ")