Many dialects show the stem *mōn- in the meaning 'son' (and *mōḷ- in the
meaning 'daughter'). These forms certainly go back to *magan and *magaḷ
respectively; however, the contracted forms should probably be reconstructed
for PSDR, as they are represented not only in Tamil/Malayalam (where
intervocalic -g- regularly disappears in some dialects), but in other languages
as well, most notably Tulu and Toda.
Another irregularity is the development *mag- > mog- that happens in various
languages; particularly noteworthy is the development in Kota, where *-a- does
not normally lead to -o- in such a context. Since the root's usage was
particularly frequent, special irregularities cannot be excluded; however, it
is not excluded that we have to deal with a special cluster (maybe even a
special phoneme) in this case.

    The same root was probably used in PA to denote 'small finger' (cf. the meanings in PTs) later developped to 'finger' in some languages: Tind. miḳ́a, Akhv. ũḳa, Tseg. ĩḳʷa ( < *ʔi-miḳV with a laryngeal prefix occurring also in some other body part names), Cham. miḳ- in miḳ-s:il 'finger-ring' ( = Tind. miḳ́i-š:al).