ab o:uo:

Slavic *aje, Germanic *ajja-, Latin o:uum, (Att.) Greek o:Io'n (Arg.
pl. o:'bea), Iranian *a:ya- 'egg', Armenian ju are usually thought to
continue PIE *o(:)w(y)om, which in turn is explained as vrddhi of the
'bird'-word ('of bird'<'bird') reflexed in Latin auis 'bird'. But the
latter rather points to something like *h2ewi- as a proto-form, and I
can't see how a least common denominator for *o(:)w(y)om and *h2ewi-
could be found, taking into account the fact that some branches
doesn't demonstrate any evidence for *-w-, while others for *-y-, and
Germanic doesn't show any vrddhi at all.

Any ideas?

Sergei