just verifying a point

Some romanic linguists see that the latin equa became "ebba"
in saridc and "iapa" in romanian ; so , there they like to
observ a labialisation of latin "qu".
I ask myself why these rule forget something else. And in this
case that in P-celtic dialects we have an teonim Epona = the
goodes of horses and in celtic "epo" meant "horse, and in the
same time we have too thracians antroponims like Eppa.
Because someone doubt gratefully about this I should like to
see your opinion about that. The fact of a pre-latin
labialisation of "kW" in thracian and celtic could too be
sustained with teh PIE *akWa which got in romanian "apa" how
the dacian toponims Salapia , Naparis, etc. let us know.