________________
JULY, 1882.]
tively by the facts derived from a comparison of the languages of Magadha and Ceylon. Various dialects have been assigned to Magadha. Pâli we may no longer take into consideration as its different origin has been proved. (Westergaard and Kuhn take Pâli to be the language of Ujjeni, but Oldenberg (Vinayapitaka, Introd. p. 54) thinks that its original home was in the kingdoms of Andhra and Kâlinga.) There remain three dialects which bear the name of Magadha, viz. the dialects in which some of Aśoka's edicts are written, the dramatic Mâ gadhi, and the language of the sacred books of the Jaina sect.
CONTRIBUTIONS TO SIMHALESE GRAMMAR.
The investigation of the dramatic Magadhi has presented no slight difficulties, on account of the apparent preservation of an ancient phonetic condition long left behind by contemporary dialects and even by the dialects spoken in the time of Asoka. I allude here to the substitution of st for tt and shth (Hemachandra, IV, 289) and of st for sth and rth (Hemachandra, IV, 290). If we had only st for shth, and st for sth, the conclusion would be most natural that these combinations were the immediate successors of the corresponding ones found in Sanskrit, but that s also stands for !!, and st for rth alters the case. I believe the Skt. shth had first changed, as in all Prâkrits, into tth, and subsequently lost the aspiration, rth into tth and tt. It was to avoid the difficult pronunciation of a double consonant that, in Magadhi, the first of them was changed into a sibilant, and thus titta tirtha becomes tista, conf. nirastiya 'disinterested' (inscr. of Shahbâzgarhi, in Cunningham, Reports, vol. I. p. 78). Atta, artha asta, sustu; suṭṭide, patta pasta, suttu, sushṭhu susthita sustide, &c. It is not easy to account for sk ksh in preksh and áchaksh, Hem. IV, 297, but as this change is limited to these two verbs and the precedent of a sibilant being put before a hard consonant was given in the case of
Aboka.
nom. e loc. si
tuphe (Kern Aśoka 102) puluva (Cunn. I, 69) pulva
Dramatic. do.
do.
do.
purawa (Hem. IV, 323) in Sauraseni kalia (Hem. IV, 302) kadua (Hem. IV, 272)
dentals, it may very well also be regarded as secondary. ich: chh (Hem. IV, 295) is the third remarkable change, but this cannot be old. The possibility remains to account for all these changes by an antipathy to aspiration, which would have become hardened and put before the consonant; but then there is tṭ which has no aspiration treated in the same way.
Quite different from this artificial language is the Mâ gadhi of the Jain a sect, or, as it is now generally called, the Jaina Prâkṛit. The only point in which both agree is the termination e of the nom. sing. in the first declension (Hem. IV, 287), and this termination is also to be found in Aśoka's inscriptions (not only in the nom. sing. but wherever the corresponding Sanskrit form terminates in as, for instance lajine = rájñaḥ, lájáne rájánaḥ and in Simhalese), whereas the Pâli preserves the ancient o of the Sanskrit. The principal peculiarities of the Jaina Prakrit are the change of a single consonant in the middle of a word into y (Hem. I, 177) which is the last stage before the dropping of the consonant as found in other Prakrits, the change of initial y to j and of yy to jj, the change of aspirates to h. (See my Beitrage zur Grammatik des Jaina-prákrit, p. 12.)
The name of Ardha magadhi, by which the Prakrit grammarians call this language, does not help us in finding out its position amongst the Prakrits, for it was applied to different dialects at different periods, as may be seen from Hem. IV, 287, Comment. We can however fix its position between the Pâli and Magadhi of the inscriptions on one side and the Maharashtri on the other.
-
Mágadhi.
I now proceed to give a small comparative table of Magadhi and Simhalese words and grammatical forms, and then to point out some such difference as must have existed between the two languages before the departure of the Simhalese from India.
Jaina. do.
msi
tubbha, tubbhe (Hem. III, 90) pure kamma (Hem. I, 57)
pagijjhiya (Hem. IV, 216) kattu (Hem. II, 164)
199
Simhalese. do.
hi topi pera, pura
kariya kotu
