194 Proceedings bfthe Asiatic Society. [June, 



Panjabi and Sindhi. Gujarati hovers between the two. It is possi- 

 ble to construct a long sentence, nay to write a book even, in Hindi, 

 Persian, Arabic, and Turkish, in which sixty per cent, of the words 

 used should be identical, because borrowed from Arabic ; yet no one 

 woidd conclude that these languages were connected. Similarly a 

 book may be written in Bengali, Uriya, and Marathi, with the same 

 proportion of identical words, and yet no argument could be thence 

 derived for or against the connection of the languages. 



The fact is that the Sanskrit words so largely employed by pandits 

 in Bengal and Orissa, are not living words at all, they are dead, 

 dead ages ago, and only now galvanized into the semblance of life ; 

 they form no part of the real working stock of words of the lan- 

 guage. When they died ages ago, their sons inherited their place, 

 and now their grandsons or great-grandsons hold it. In plain Eng- 

 lish, such Sanskrit words as were used by the Uriyas and Bengalis 

 twenty-five centuries ago, have since then undergone the usual fate 

 of words, and have been corrupted, abraded, and distorted, till they 

 often bear no resemblance at all to the original word. And it is these 

 corrupted, or as they are called Tadbhava words, that are the real 

 living words of the language, the words that have worn into their 

 present shape by long use in the mouths of the people. These 

 words our fastidious writers reject, and when by going back to the 

 Sanskrit for their words, they have composed a work to their taste, 

 lo ! they say Uriya and Bengali are one language ; for proof, read 

 such and such works. I would suggest rather, let them take a chdsa 

 of Dacca and a chdsa of Guinsar, and see how much they understand 

 of one another's talk. 



In the grammatical structure of the Uriya language, we see 

 traces of a very well defined Prakrit with features peculiar to 

 itself. I begin with the verb as the simplest part of the language, 

 (in this case at least). 



There is first a present participle in w, as chain, and a past parti- 

 ciple in i f as chali ; by means of these two a whole string of com- 

 pound tenses are formed, thus — 



( aclihi I am going 

 I. chalu j thili I was going 



^ hebi I shall be going 



