Sandhis of word-final N

[Madhav Deshpande]

Another feature I have noticed in the execution of this particular sandhi
in its Paninian/commentarial usage is that the doubling of the final
dental n is followed more consistently, e.g. san+anta > sannanta.
However, the final .n and G (velar nasal) are not seen doubled, e.g. iko
ya.n aci and not iko ya.n.n aci in the reading of the Astadhyayi, and
tiG+anta > tiGanta and not tiGGant, except in the reading of P.8.1.28
(tiGG atiGaH).  But again the expressions like tiGanta and yaGanta are
more commonplace.  Within this irregularity, the doubling of the final n
is more or less regular.  Perhaps this has to do with the frequencies with
which these consonants occur in the word-final position in normal
Sanskrit.  Sorry about the mixed notation above.
                                        Madhav Deshpande