other times in that of Sansk. na. An analogous case is Sansk. atho, almost imperceptibly differing from atha. Perhaps the most curious of similar forms in the Gathas is co, in meaning exactly coinciding with ca; this co I take to be the older form of the Magadhï cu in the Agoka edicts.
From the occurrence of peculiar old words and forms we may draw inferences as to the age of certain compositions in ordinary cases; but it is not safe to apply the same test, if there is sufficiënt reason to suppose that the work, the date of which we wish to determine, has been carefully moulded upon time-honoured models. In such a case new words prove a good dealold ones next to nothing. Therefore it would be an abuse of the argument ex silentio to infer from the total absence of such new words in our Saddharma-pundarïka that the bulk of the Sütra must date from the earlier period of Buddhism.
I had already occasion to notice that the two versions, the prose and the metrical one, in our Sütra show here and there material discrepancies. The question arises to which of the two we must award the palm of prioritj. Repeatedly, both in prose and poetry, the Sütra is spoken of as consisting of stanzas; e. g. chap. vii, st. 82; chapters x and xxii in the prose portion, several times. As the term of stanza (gatha), for aught I know, is never used to denote a certain number of syllables, there is astrongpresumption that the ancient text consisted of verses, with an admixture of short prose passages serving as introduction or to connect the more solemn poetical pieces. The idea to expand such passages into a regular prose version would especially recommend itself at a period when the poetical dialect began to become obsolete and obscure. Without being a formal commentan , the prose version would yet tend to elucidate the older holy text.
It will not be objected that, because not all chapters in the Saddharmapundarïka have a poetical version added, the original cannot have been a poem. For the chapters containing but one version, viz. xxi, xxii, xxiii, xxv, and xxvi, show decided tracés of being later additions; and as to the final chapter, it may be held to be a moderate amplification of a short prose epilogue.
In contending that the original text of our Sütra was probably, in the main, a work in metrical form, I do not mean to say that the poetical version in all the chapters must be considered to be prior to the prose. - The Gathas of the Saddharma-pundarïka are nowhere very brilliant, but in some
1 As e. g. the word dïnara in the Aijoka Avadana; the passage on the Greeks Youas, in Assalayana Sutta (ed. Pischel), p. 10; cf. the editor's remark, p. 6; the word karama for kalama, calamus to write with, in Karanda-vyüha (Calo. ed.), p. 69.
» Isolated stanzas, as in chapters xxii, m, and elsewhere, are wholly left out o question.