Gk gen plural

[petegray]

Does anyone happen to know if there is any actual evidence for the genitive
plural ending -a:so:m in Greek, rather than the normal -a:o:n and its
derivatives?

If not, what basis is there for the assumption that this ending was imported
from the pronouns?  Can't the attested forms be derived (with the help of
paradigmatic analogy) from an original -ah2-om?  Mycenaean seems to show no
sign of the -s-.

Peter