The general argument of Walleser ( Sprache und Heimat des Pāli-Kanons ) is as follows: The agreement of all old sources that the language in which the Tipiṭaka was brought to Ceylon was that of Māgadha is too strong to be lightly cast aside. The capital of Magadha was Pāṭaliputta (Sanskrit Pāṭaliputra). In Buddha's time it was known as Pāṭali-gāma Tāṭali Village.‘ Walleser accordingly goes back to Wilson's theory that Pāṭaliputra is a transformation of Pāṭalipura Tāṭali Fortified Town.‘ The language therefore would be known as Pāṭalī Bhāṣā, or simply Pāṭalī, in the same way that the language of Magadha was known as Māgadhī. Now it is well-known that in the Middle Indo-Aryan dialects stop consonants (when unaspirated) have a marked tendency to drop out, and frequently the vowels thus brought into hiatus contract. Examples of this are found even in Pāli, thus Kusināra (Sanskrit Kuśinagara). So Pāṭali could become Pāali and finally Pāli. As a parallel on the Asokan inscriptions Peteṇika (so!), from *Prātiṣṭhānika (a vṛddhi derivative of Pratiṣṭhāna) is adduced. Hence Pāli is the language of Pāṭaliputta (Sanskrit Pāṭaliputra). The charge that Pāli can not have had its home in Magadha because of its phonetic and grammatical peculiarities can be met because this does not take into consideration the wide territories of Magadha: within these a rather considerable range of dialectic peculiarities is possible; again, Pāli as a literary language existed over too long a stretch of time to expect uniformity in every detail; Pāli being contemporaneous with Buddha is naturally more archaic than the Asokan dialects, but in any case Pāli is near the dialects of the eastern Asokan inscriptions.—A wealth of literary references are given by Walleser to support his views, and a number of other linguistic points are treated incidentally.
T. Michelson (Fri,) studied this question.