The History of Indian Literature c8.337

c8.337

∗ In the latter case a difficulty is caused by the *Káṇva* recension, which subdivides the last *káṇḍa* into two parts (xvi., xvii.); this division, however, seems not to have been generally received, since in the MSS. of Śaṃkara's commentary, at least, the Upanishad (xvii.) is reckoned throughout as beginning with the third *adhyáya* (viz., of the *káṇḍa*), so that xvi. and xvii. coincide.—[A highly remarkable statement is found in the MSS. of the Mádhyaṃdina recension at v. 3. 1. 14, to the effect that this point marks not only *káṇḍasyá 'rdham*, with 236 *kaṇḍikás*, but also, according to a marginal gloss, *śatapathasyá 'rdham*, with 3129 *kaṇḍikás*; see p. 497 of my edition. As a matter of fact, the preceding *kaṇḍikás* do amount to this latter number; but if we fix it as the norm for the second half, we are only brought down to xii. 7. 3. 18, that is, not even to the close of the twelfth book! The point which marks the exact half for the present extent of the work (3812 *k.*) is at vi. 7. 1. 19, where also the MSS. repeat the above statement (p. 555).—It deserves special mention that the notation of the accents operates beyond the limits of the individual *kaṇḍikás*, the accent at the end of a *kaṇḍiká* being modified by the accent of the first word of the next *kaṇḍiká*. From this we might perhaps conclude that the marking of the accents is earlier in date than the division of the text into *kaṇḍikás*. As, however, we find exactly the same state of things with regard to the final and initial words of the individual *bráhmaṇas* (see *Jenaer Literaturzeitung*, 1875, p. 314), we should also have to refer the *bráhmaṇa* division to a later date, and this is hardly possible].