________________
MARCH, 1880.]
MISCELLANEA
81
the exception, therefore, of Indra, more perhaps is an and the augment num, intervene between the known of Chandra from quotations in existing ror sh and the dental 1. Chandra, on the other works than of any one of the other three whose hand, after giving a rule corresponding to Panini works have shared the same fate-viz., Kabakrits- viii. 4, 1, in nearly the same words, mentions in a na, Apisali, and Amara. What is known of him, subsequent sätra a number of letters the intervenhowever, is so scanty that it would not enable us | tion of which between the r or sh and the dental to form any definite idea of the character and n prevents the operation of the first mentioned extent of his work. To say that it consisted of rule. These letters are those belonging to the Batras, and that it was quoted by Bhattoji in the palatal, cerebral, and dental classes of consonants, Praudhamanoramd, by Ujjvaladatta in his com those included in sar and the consonant l. The mentary on the Uņádi-sútras, and by one or two obvious meaning of this is that the intervention other authors, would embrace perhaps all that of any other letters and the only others are those could have hitherto been said of the Chandra given by Påņini-does not prevent the cerebral grammar. But the MS. just discovered enables change. us to ascertain with certainty several important At also occurs in Pånini viii: 4, 63, but in the facts regarding this work and its relation to the corresponding sutra of Chandra we find, instead of existing systems of Sanskrit grammar.
it, the more comprehensive pratyahâra am, which From this MS., of which I shall give & short includes, besides the letters contained in at, the account before concluding this paper, we gather nasals and the consonant l. That Chandra's rule is that Chandra's work wae based on the model of the better of the two will be readily admitted when and was intended as an improvement on, Pånini's it is remembered that Pånini's sätra was amended Ashtadhydvf. Both works hare adopted the by a vârttika of Kåtyåyana substituting this very same arrangement, which may be termed arti- pratyåhåra am for the less comprehensive at. ficial, as distinguished from the natural arrange- Another praty&håra employed by Påņini but ment according to subjects observable in the dispensed with by Chandra is the an formed with Katantra and Mugdhabudha, and, in fact, in all the second in the scheme. This occurs only European treatises on grammar. The division in one solitary sátra of Pånini, viz., I. i. 69, accordinto adhydyas and pddas is also common to both ing to which the letters contained in it would also works, with only this difference, that the num- represent or imply their homogeneous modificaber of adhydyas in Panini is eight, while that in tions; for instance, the short vowel a, which alone Chandra is six, This has resultod probably from is given in the scheme, would imply, among the latter embodying in one adhydya the sub- others, the long and prolated d. Chandra finds no stance of Pånini's third, fourth, and fifth chapters, necessity for this role, and omits it altogether, as all of which treat of affixes.
the short vowel a, being a játi term, would necesBoth works open with a peculiar arrangement sarily imply its long and prolated forms and its of the letters, materially differing from the alpha- other homogeneous variations. betical arrangement, and styled by the Paniniya The only other pratyâhâra employed by Panini school of grammarians the Sivasútras. There is, and omitted by Chandra is jhas ; but he employs however, a slight difference between the two in his grammar the four pratyaharas, rik, ham, schemes-viz., that the it or anubandha t at the mam, and chay, which do not occur in Påņini. end of the fifth sutra in Påņini's is omitted in With reference to the last of these, chay, it Chandra's scheme, which thus contains thirteen must be stated that, though not employed by sâtras, while the number in Pâqini's is fourteen. Pånini, it is used by Katyayana in a vârttika to This omission causes the absence from Chandra's viii. 4, 48. work of only one pratyahara, at, the necessity for With the aid of these new pratyaháras Chandra which is obviated by Chandra in an ingenious has effected a decided improvement on many of manner, as will be seen from an instance which I the satras of Påņini, an instance of which may cannot help adverting to.
here be mentioned. The sutra Påņini i. 1, 51Panini, after laying down in viii. 4, 1, that the Uran raparah-means, if we assign to the terms dental n should be changed to the cerebral when occurring in it their true and legitimate significait follows r or sh in the same pada, states in the tions, that an, that is to say the vowels a, i, and sutra immediately following that the rule obtains 16, substituted for the vowel ri, should be followed even when the letters included in the pratyábåra by r. It was found by grammarians that this at, and those belonging to the guttural and labial rule stated too little, as it did not teach that an, classes of consonants, as well as the preposition substituted for the vowel li," should be followed
Almost all Oriental scholars transliterate this vowel retaining only u, but they are too long to be stated in a by "lri." I have my reasons for omitting the and foot-note. Soe Max Muller's Sanskrit Grammar. p. 18.
