________________
252
THE INDIAN ANTIQUARY.
[SEPTEMBER, 1894.
certain proportion of the roots is the result of misreadings. This is, of course, highly probable in all cases where the Dhaga patha gives pairs like go and go or ag and 39. 'T'he characters for jha and ú are almost exactly alike in the Nagari alphabet of the ninth, tenth, eleventh and twelfth centuries, just as those for ya and pa in the later MSS. More important is another point, which likewise has been frequently noticed, vis., the fact that only a small portion of the Vedic literature, known to Påņini and his predecessors, has been preserved, and that of the ancient laukika Sastra, the Kavya, Porâņa, Itibisa and the technical treatises only very small remnants have come down to our times. The assertion that the old literature has suffered terrible losses, is admitted by all Sanskritists. It is only a pity that their extent has not been ascertained, at least approximately, by the preparation of a list of works and authors mentioned in the 'Sabdannsasana, the Brahmaņas, the Upanishads and the Veddityas. Such a list, especially if supplemented by an enumeration of the numerous references to the spoken language, which Panini's Sabdanusisana contains, would probably bar for the future the inference that a root or fortu must be fictitious, because it is not found in the accessible literature. This inference is based on a conclusio a minori ad majus, which with a list, sbewt ing what existed formerly and what we have now, would at once become apparent. The loss Såkbas of the Vedas and the lost works of the laukika Sastra amount to hundreds. If on an average a third or a fourth of them contained each, as is perhaps not improbable according to the results of the exploration of recently recovered Samhitas and Sutras, one or two of the as yet untraceable roots, that would be sufficient to account for all the lost stems.
Three other considerations, it seems to me, help to explain some of the most remarkable peculiarities, observable in the materials incorporated in the Dhâtupâţha, viz., the fact that a certain proportion of the roots really is and will remain isolated, neither derivatives nor cognate forms being traceable in the Indo-Aryan or in the Indo-European languages, and the indisputable fact that many roots may readily be arranged in groaps, similar in sound and identical in meaning and inflexion. Both these peculiarities, as stated above, have been used by Prof. Edgren in order to prove that the verbs, shewing them, must be fictitious. And it has been pointed out, that the number of the isolated and barren verbs is not so great as Prof. Edgren supposes, the inflected forms or representatives of a certain proportion being found in the Prakrits and in the unexplored Sanskrit literature. Nevertheless, & certain number of instances will remain, which requires accounting for. With respect to the second fact, it has been pointed out that many of the curious variants are clearly dialectic and derived from lost or preserved parent-stems in accordance with phonetic laws valid in the Prakrits and in Sanskrit.
The chief considerations, which in my opinion do account for these peculiarities are (1) the yrent length of the period, during which the materials of the Dhâtupâțha were collected, (2) the enormous extent of the territory from which the Hindu grammarians drew their inguistic facts, and (3) the great diversity of the several sections of the Indo-Aryans inhabiting this territory
It is admitted at all hands that Pånini's Sabdanuáasana is the last link in a long chain of grammatical treatises, which were gradually enlarged and made more and more intricate, until the Hindu system of grammar became a science, which can be mastered only by a diligent study continued for years. According to the unanimous tradition of the Hindus, the Vyakaraņa, is a Vedâúga, i. e., a science subservient to the study of the Veda, and it is highly probable that the older Hindu grammars exclusively or chiefly explained the Vedic forms, just like the oldest Koslas, the Nighantus, include very little that is not derived from Vedic texts. In Papini's grammar the Vedic language is of minor importance. Its chief aim is to teach the correct
A perusal of Prof. Per Person's Wurzelerweiterung und Wurzelvariation would perhaps convince Prof. Edgren that many Indo-European roots may be arranged in ganas, similar to those in which he has arranged so many Terbs of the Dhátupåtha.
