ratyaa, ratya.m

Dear Friends,

I've entered into study of Pali the Thai way, surrounded by midget
samenaras yelling at the top of their lungs and running circles around
me in their ability to memorize Pali texts. I have a question regarding
the grammar we are studying. It says that under the paradigm for "ratti
(f.)", 5th and 7th vibhatti, ekavacana can become "ratyaa" and "ratya.m"
respectively. Since ratti is the word used as an example, it appears
that they would then have us conjugate "nandi" as "nanyaa, nanya.m" and
"vati" as "vayaa, vaya.m" !? Reading through Duroiselle's grammar, it
gives the same impression:

     (d) Before ā, of the same cases, final i of the stem may become y by
     rule 27(i)-a; and as in Pāli there can be no group of three
     consonants* one t is dropped. Hence we get: ratti + ā = rattyā = ratyā.
     *Except ntr, as in antra, etc.


Our teacher, Mahavituun explained it clearer, saying that in his
understanding, the rule should be "where three consonants are grouped,
*and two of them are the same*, remove one of the two that are the
same." This makes it clear, but then how to conjugate "nandi"? Should it
be "nandyaa" or should it be assumed that this form simply cannot exist
(according to Maha Vituun)? I am wondering whether there is any
clarification regarding this in the ancient grammars.

Thanks,

Yuttadhammo