________________
208
THE INDIAN ANTIQUARY.
6; Vârt. 1 and 2 on P. VI. 1, 58; Vârt. 1 on P. VI. 1, 108; Vârt. 1 on P. VI. 1, 171.
On P. V. 4, 103,
refer, I have given as a remark of Patanjali's the statement भनसन्तान्नपुंसकाच्छन्दसि वेति वक्तव्यम्; the MS. 2. omits here अनसन्तान्नपुंसकाच्छन्दसि and g., a very indifferent MS., has the same words twice. The repetition of Panini's rule at the beginning of this statement must again make us suspect that a Vârttika has disappeared in the MSS., and that we ought to read - पुंसकाच्छन्दसि वा ||१|| अनसन्तान्नपुंसकाच्छन्दसि वेोते . It is true that Kâtyâyana, when he wishes to make a rule of Panini's optional, generally employs the word (compare e.g. Vârt. 1 on P. III. 1, 27, Vârt. 1 on P. V. 1, 10, Vârt. 1 on P. VIII. 2, 103), yet on P. III. 3, 156 हेतुहेतुमतोर्लिङ्, we have the Varttika हेतुहेतुमतोर्लिङ्गा, to which अनसन्तान्नपुंसकाच्छन्दसि
would be similar in every respect.
The doubts and conjectures, which I have expressed in the above, are in the first instance suggested by the principle that the first Vårttika on a rule must contain some distinct reference to that rule. There are other matters, on which I should wish to elicit the opinions of Indian scholars.
4
We frequently find in the Mahabhashya statements, which end with the word, 'something has been said,' 'a remark has been made. They are invariably followed by fang what has been said ?' When we examine the answers to this question, we are struck by the fact, that almost everywhere they con. sist in one or more Vârttikas met with in different parts of the Mahâbhishya. We are thus led to the conclusion that the statements ending with 3 are themselves Varttikas, and that in them Katyayana is referring us to other Vârttikas, which generally precede, sometimes follow, the Vârttikas ending with उक्तम्. And our belief in the soundness of this conclusion is strengthened, when we examine those statements a little more closely.
When we try to ascertain what Vârttikas Katyayana may have appended e.g. to P. I. 1, 60, अदर्शनं लोपः, the first Vārttika to present itself will be rearracing. It is explained by Patanjali as other Vârttikas are, with this difference only that the comment on the words preceding has taken the form of an introductory remark. It does contain
[JULY, 1886.
the necessary reference to Pânini's rule, for like other Vârttikas on Samhjá-rules it does begin with the technical term defined by Panini, compounded with arr. It must be considered a Vârttika, because the wording of the sentence which in my edition is given as Vârt. 2, and which undoubtedly is a Vârttika, shows that there must be a Vârttika preceding it, and because in the preceding we find nothing that could possibly be regarded as one, excepting our लोपसंज्ञायामर्थसतोरुक्तम्. “The remarks that have been made," and to which we are referred by उक्तम्, are इतिकरणोऽर्थनिर्देशार्थः and सिद्धं तु नित्यशब्दत्वात्, the former being the Vârt. 3 on P. I. 1, 44, and the latter the Vårt. 9 on P. I. 1, 1.
Now, what I have proved in this one instance may be shown to be true elsewhere, and it be comes evident, that by the Vârttikas described Katyayana has furnished a means of testing to some extent the value of any attempt at reconstructing his work. For whenever we find a statement of his ending with 3, our Varttika-pátha, if it be right, must contain the Vârttika or Vârttikas, to which he refers us. And when, to give a somewhat striking instance, in the Vârt. 10 on P. I. 2, 45 he tells us "that he has said something regarding the question as to whether letters have a meaning or not," and when, in order to show what Katyayana has said, Patanjali quotes the seven Vârttikas on pp. 30 and 31 of Vol. I. of my edition, exactly as they have been printed there, it will appear probable that the principles followed in the reconstruction of the Vârttikas were correct. On the other hand, the question किमुक्तम् should everywhere be answered by Varttikas only; and if anywhere in the Mahâ. bhashya it should have been answered differently, it must be possible to substitute a Vârttika or Vârttikas for the answer actually given by Patanjali. If in any particular case it should be found impossible to point out the Vârttika to which Kâtyâyana could have referred us, such a case should either make us doubt our having fully understood the drift of his remarks, or suspect some fault in our method, or in the readings of the MSS, used for the reconstruction of his work,
In my edition the number of Vârttikas ending with is 45; not one of them has been given in the Calcutta edition of Pânini. 38