________________
380
THE INDIAN ANTIQUARY.
[SEPTEMBER, 1903.
corresponding passage in Panini, the Mahd. the classes is, however, different, Chandra bhdshya, or the Kddiká, when such exist. The grouping them according to the voices of the whole work is divided into six books (each sub- verb. Thus in the second or ad-adi class we have divided into four oddas), which correspond to the first 36 Parismaipade verbs (called atanand), last six of Påņini, the contents of the latter's then 19 Atmanêpada ones (taidninah), and lastly first two books being scattered over various parts seven conjugated in both voices (vibhúshitál). of Obandra's grammar. The first thirteen stras. Prof. Liebich points out the interesting fact Àre identical with Panini's fourteen Siva Satras, that the Dhatupatha of the Katantra Gramthe only variation being, that the fifth and sixth of mar is in reality that of the Chandra system Panini - hayavarat lan-are here run into one : 1 as modified by Dargasimha, the well-known comhayavaralan. The degree of correspondence mentator of the Kdtantra. This modified between Chandra's sätras and Påņini's varies. Chandra Dhatupátha is No. 3727 in the Tanjur.
There is sometimes complete identity; e. 9., The genuine Dhatupátha of the Katantra is also upamdndd achdrd (I, i, 25 = P. III, i, 10).
preserved in the Tibetan version as No. 3723 in Occasionally the sutras differ in the order of the the same encyclopædic collection, but the original words only, as bid anékal sarvasya (I, i, 12) =
Sanskrit text appears to have been lost. anekal kit sarvasya (P. I, i, 55). The agreement
Prof. Liebich bas constituted the text of in many cases is only partial; for instance,
the Unddi-sútra from two incomplete Nepalese shashthydntyasya (I, i, 10) = alt 'ntyasya (P.I,
Sanskrit MSS. of the Unda stitra-vritti. One of i, 52). The identity is, lastly, often restricted to
these MSS. contains, in this commentary, all the the matter, as riks 'no ralau (I, i, 15) = ur an
Satras, except the last fifteen, of the first two raparah (P. I, i, 51). Chandra's new material,
books, while the second MS. has preserved the amounting apparently to about 35 Satras, has
whole of the third book except the first six Satras. for the most part been incorporated in the
Fortunately the missing twenty-one Satras can Katika.
be restored with certainty from the very faithful The commentary on this Satra, the Chandra- Tibetan version of the Tanjur vritti already referred to, is no doubt the most
Chandragðmin disposes the Unddi words in his important text of the whole system. Oorre
three books independently of Panini, the suffixes sponding in style and treatment of the subject
being arranged by him according to their final matter to the Kábika Vritti, it will most probably
letter; he also frequently derives the words in render important help in emending the corrupt
a different way. Owing to the great divergencies passages of the printed text of the latter commentary. It is therefore to be hoped that Prof.
in this case, Prof. Liebich has not thought it
worth while to refer to the parallel edtras of the Liebich will before long publish the complete text of
Paninean system, He has, instead, added after this valuable work also. In the meantime, he has
each satra the complete Unadi word and its in the present volume given extracts from it, in the
paraphrase from form of notes to those portions of Chandra's atra
the commentary. Thus the
Satra III, 105, pabeh mut cha, is followed by the which have no parallel in the Påninean system.
derived and explanatory words vakshah: krodah. Prof. Liebich's text of the Dhátupatha-sútra is
The Tibetan version retains the Sanskrit synonym based on a Cambridge MS. (dated A.D. 1356) and
intact, but adds to it the Tibetan translation. on one of the three recensions of the work in the Tibetan Tanjur (No. 3724). That recension he
Prof. Liebich is to be congratulated on his regards as containing the purest form of the thoroughly accurate and scholarly publication, Dhatupatha of the Chåndra system (see Gött. which not only marks in itself an advance in our Nachrichten, p. 804). The number of roots knowledge of the intricate systems of native enumerated in that recension is 1,659, while that grammar, but is also a stepping-stone to further in Liebich's text is, according to my reckoning,
advance in the subject. Dis work should nlso be hardly 1,600. It would be interesting to have this
specially welcomed by Indian Sanskritiste, to discrepancy explained. The Satras are here whom the study of the native grammar is printed in two columns to the page, the number a peculiarly interesting and important branch of of the corresponding root in Böhtlingk's Sanskrit learning. The employment of German Dhatupatha contained in his edition of Panini, being almost entirely limited to the brief preface, 1887) being added in each case. In matter, a knowledge of that language can be dispensed Chandra agrees here pretty closely with P&ņini, with for the perusal of the book. the roots being similarly divided according to the
A. A. MACDONELL. ten classes. The arrangement of the verbs within
Oxford, March 10th, 1903.