________________
186
variant turning up in the South-Indian MSS of Pkt texts or in Pkt text's written in South-India. See, for instance, Karpūr amañjart I 7, MSS S and U, which have a variant pāuça, and MS T, which has yāțaya, and Candralekha I 11 and Kamsavaho IV 48h. The latter two texts were written by South-Indians, Rudradāsa and Rāma Pāṇivāda respectively. Hence pāada could be an innovation in the South Indian MSS of the Sattasat. On the other hand this does not explain how pāada could have become prevalent in the South and how the notion that it is the correct form could have arisen there. In this connection it should be noted that pāada, from prākrta, is formed after a sound-change, -rta- + -ada, typical of AMg. and JM. (Pischel S 219). Its occurrence in the South-Indian MSS of the Sattasat could then be explained with reference to the Jaina. Nāgart MS from which they ultimately derive. The word may have been introduced in this MS by its scribe who very likely was familiar with AMg. and JM. or with the sound-changes accounting for the derivation of these dialects from Skt. Furthermore, from its occurrence precisely in these South Indian MSS of the Sattas as it may have become known in the South as the only correct form for Skt prāksta. However, M. Pkt usually has pāua. Also the earliest AMg. instances of the word, found in late canonical texts, which, however, are considered in any case, older than the Sattas at, read pāgaya or pāyaya, showing a sound-change similar to that in pāua. The word pāada therefore most likely seems to be a learned, hypercorrect Alg. form.
It may not be without significance that pāada, which is thus well known in South India, is not mentioned in any of the grammars (pauda in the pratyādigana in Prakrtasarvasva II 10 is a corruption, through paduga (MSS U, I, O and G), for pāhuda (cf. Hem. I 206)). This would indicate that the grammars did not use or know the South Indian MSS of the Sattasaf.
After na the verb jāņa- (Skt jñā-) invariably loses its initial j-; see, for instance, 347 na āņimo. This loss is usually restricted to -j- found within a word. Apparently the phrase na āņa-, which occurs frequently, has come to be considered as one word (cf. na ittho in 502). All the MSS, however, except Ma and y, have na jānanti here, which nevertheless is to be rejected as the easier reading.
