________________
174
THE INDIAN ANTIQUARY
[NOVEMBER, 1916
there. Its omission is strange and ur.accountable. Perhaps it is the scribe's mistake, who, seeing that all other statements of Gârgya are supported by examples, wanted to connect the examples oft: T: with the first statement. The original sense of afara a being obscure to him, he appears to have understood it as Durga understood it later and striking off a TETET , connected it (i. e. F TATA etc.) with the sentence ending in Far A.
Max Müller has a different construction. He makes the first sentence end with tra, taking these to be examples of the case where Gárgya and the Nirukta's agree. saman
- would be in themselves intelligible'. To Gârgya however pt: 8: etc. are not examples of regular formation, as his objections show. See Max Müller Ane Sansk. Lit. 165.
IV R. 39. 11ff. S. II 137, 18ff fooz T a ra Ara | For Pargaret 903 7............... Hafa fra
egetat i atca
Here the difficulty is caused by the one compound देवतानामप्राधान्येन. Durga s. 532_has अथ पुनर्यत्र 'नैषण्टक' T E STATY9 eto, as explanation. This is called egen owing to the subordinate nature of gods'. In the first place this way of interpreting the phrase makes the following line turata etc. (that, which falls in a verse dedicated to another god is om ) quite redundant as the same meaning is apparently briefly expressed by the phrase in question. Secondly, this way of taking the passage does not do full justice to the two T. On the very face of it, the passage offers two words or names that re so to say pitted against each other by the parallel expression T . Thirdly this sort of explanation ignores the force and the propriety of the parallel phrases introduced by or etc. and mana EITT. They are explanations of the two classes of words that are mentioned in the head line and that the author is anxious to define and distinguish clearly.
Roth has not got any note on the passage. There is however an indication in his Einleitung P. XIII, that he took the passage to mean this is tough owing to the prominence of the names of gods'. He has given a general idea of the whole passage beginning from ET Tranha E TNE. The translation of the closing portion, which only is pertinent here, runs thus:- The following generations, then, composed this book also in which are enumerated, the roots for one activity, the nouns for one idea, also words that have several meanings and lastly the names of gods.' The last line suggests that he understands the passage as just indicated. If so the V-19 and the parallel expressions which appear to be purposely put to distinguish between two kinds of names viz., OT and quia arra etc., are not well explained. The following is I think the proper way of explaining the pasasge.
We have first to separate the words 99 and grada. The passage then reads Agora ATA, Tyruar This name of a god is on this one (however) is primary.' Having first of all postulated two kinds of names for gods, he proceeds to explain them in turn. U names are those that occur in a verse for another god; while those that contain the praise of certain gods primarily i. e. without being subordinately mentioned with others) are ta names. The word aga then gets an extensive application. It means then, not only subordinate names of gods but in a general way, such other names as occur in verses in praise of a particular god. An example of a name is t r y where st is to because it occurs in a verse for another god. See
R. 49. 11 Taaret Hoca swarar a i. e. these (rivers) are very often secondarily mentioned but rarely primarily,' नैघण्टुकं वृत्तं is a synonym of निपात e.g. R. 47, 22 aparatlarda mrgrer.
