Whare did you take these "Dacian" words from? *ap-a: 'river, water' has
been "seen" in Balkan toponymy, and someting like toponymic *mariska- may have
occurred in Dacian (meaning 'marshland' rather than 'sea'), but your alleged
Dacian "mare" and "petra" look like somebody's modern figments to me. Anyway,
the Romanian words are _exactly_ what one would expect as the regular
development of Latin <petra>, <mare> and <aqua>, and any
"alternative explanation" causes extra problems without solving any.
[Moeller]
water
we have some toponimes and hidronimes with the "apa" inside . Salapia,
Naparis and more
see
we have too a toponim called Marehalis but we dont need it because mare
is a common IE so moria, maar,mira meer and so on.
stone
we have a city called Petrodava.Because in latin and romanian is almost
the same i dont see why in dacian this wouldnt be stone too. what speak against
it?
salt
i forgot to give it. The salt dacian word is too in toponims like
salapia, saline.
Piotr, I am very aware of what I try to do and how poor the material
here is. But there are a lot of words from teh substrate into romanian, belive
me. In the linguistic literature appears just the words which are common to
albanian and romanian and some another words. But really there are much more of
them.
so, the problem is, in the studiing the IE the linguists avoided for
instance linguage as romanian because this is supposed to be a neo latin
language,so , no much to do there. But i give you some examples:
PIE *bhor >traco-dac *bordelio(s) >late dacian *bordeliu >rom.
bordei,( and some others like borta, burta)
PIE *ghordos >traco-dac *gardos>rom. gard
PIE *molnos >traco-dac *malu(s)>rom. mal
PIE *legh >traco-dac *legan>rom leagan
PIE *sterp->traco-dac *sterp-u>rom. sterp
PIE *mo >traco-dac *mar-e> rom. mare
PIE *sal-,sald-> traco-dac *sal-e>rom. sare
I am sure these exemples are not enough because you have to see the
rules of treating the vowels and consonants but it maybe help a bit.