Page:A history of Sanskrit literature (1900), Macdonell, Arthur Anthony.djvu/279

 Metre, to which there are many scattered references in the Brāhmaṇas, is separately treated in a section of the Çānkhāyana Çrauta Sūtra (7, 27), in the last three sections (paṭalas) of the Rigveda Prātiçākhya, and especially in the Nidāna Sūtra, which belongs to the Sāmaveda. A part of the Chhandaḥ Sūtra of Pingala also deals with Vedic metres; but though it claims to be a Vedānga, it is in reality a late supplement, dealing chiefly with post-Vedic prosody, on which, indeed, it is the standard authority.

Finally, Kātyāyana's two Anukramaṇīs or indices, mentioned below, each contains a section, varying but slightly from the other, on Vedic metres. These sections are, however, almost identical in matter with the sixteenth paṭala of the Rigveda Prātiçākhya, and may possibly be older than the corresponding passage in the Prātiçākhya, though the latter work as a whole is doubtless anterior to the Anukramaṇī.

The Padapāṭhas show that their authors had not only made investigations as to pronunciation and Sandhi, but already knew a good deal about the grammatical analysis of words; for they separate both the parts of compounds and the prefixes of verbs, as well as certain suffixes and terminations of nouns. They had doubtless already distinguished the four parts of speech (padajātāni), though these are first mentioned by Yāska as nāman, or "noun" (including sarva-nāman, "representing all nouns" or "pronouns"), ākhyāta, "predicate," i.e. "verb"; upasarga, " supplement," i.e. "preposition"; nipāta, "incidental addition," i.e. "particle." It is perhaps to the separation of these categories that the name for grammar, vyākaraṇa, originally referred, rather than to the analysis of words. Even the Brāhmaṇas bear evidence