________________
CHAPTER FOUR
SONANTS, VOWELS AND VISARGA
SONANTS
100 Ardha-Māgadhi lacks all the sonants of Sanskrit namely F, # and e. They are represented by different vowels or syllables consisting of a vowel and a consonant.
Sanskrit ? represents the I-E. liquid , in the function of a sonant. But Sk. î has nothing to do with the I-E. *; (if it existed) which is represented in Sanskrit by -it- -ūr-. Sk. longi is the result of analogy in forms like pitin like agnin, dātīnām like agninām, Sk. I is found in one root only klp, and represents I-E. 7.
101 The vowel is changed to either 37 or zor y or R. No definite rules can be laid down for the purpose of determining the proper substitute. The usage alone decides it.
As compared to other languages Sk. alone has preserved the sonant ļ which is represented in other speeches as a combination of r and a vowel (Av. er Gr, ya, ar). The pronunciation of the sound is variously given as velar (Rg. Prāt.) alveolar (Tait. Prāt,) or cerebral (Pāņini) which may have something to do with its further development into different vowels. In actual pronunciation it is a r sound followed by a central vowel resembling i [i]. This (Cf. kȚmi krimi) distinguishes the Indian from the Iranian where the sound follows a vowel (Sk. krnoti Av. keranaoili). A trace of this nature of ? is found by Brugmann in forms like kuryāt from ks, but which are regarded by Wackernagel as analogical.
