206 Proceeding* of the Asiatic Society. [_J UNE > 



taken one example from each language, and that from books, and 

 they are not sufficient for a fair comparison of living, spoken 

 languages. What is wanted is a fall survey of the various forms of 

 the pronoun current in each province, and for that purpose a deeper 

 knowledge of the languages, both ancient and modern, and in their 

 colloquial and written forms, is required, than what I can pretend to 

 possess. As regards the Bengali and the Uriya, however, I may say 

 that in tu, tu'i, tote, tumdr, &c, there is close analogy with Bengali. 

 Amlie and tumhe, often pronounced ambhe and tumohe, are no doubt 

 peculiar ; but the change has been brought on in Bengali since its 

 separation from, or rather the birth of Uriya, and its cause is the 

 peculiar cockneyism of dropping the aspirate. 



I shall now notice the declension of nouns. Mr. Beames' sur- 

 vey leads him to the conclusion that five out of the six cases are differ- 

 ent. The very reverse, however, appears to me to be the fact. In 

 the Sanskrit, the nominative is formed in most themes by the addi- 

 tion of an s. In a sister language, the Latin, the same rule obtains 

 to a great extent, but in the derivatives of the Latin and the Sanskrit, 

 we find the mark in some cases changed to 0, and in others altoge- 

 ther omitted. In Italian and Spanish we have 0, as occhio and ojo 

 from the Latin oculus, but in the language of the Troubadours, in 

 Proven9al and in French the mark is omitted. In India, the Pun- 

 jabi and the Marwari retain the 0, but all the others drop it. The 

 result is, that the nominative is alike both in the Uriya and the 

 Bengali. 



The mark of the accusative singular in Latin and Sanskrit is m, 

 but in most of the languages derived from them, it is dropped. So 

 is the case both in Uriya and Bengali. This rule is, however, not 

 uniformly observed ; and sometimes the place of the m is supplied 

 by the syllable ku, in Uriya, and ke, in Bengali, and to trace their 

 origin, I must refer the meeting to my papers on the Gatha and the 

 Hindi dialects, where I have shown that to overcome the intricacies 

 of the Sanskrit declension, it was usual with the scalds of ancient 

 India to convert themes of various terminations to one form by affixing 

 an expletive Ic, and to mark the elision of case-affixes, the usual rule 

 \\as to add a u, which together make hu. In written Bengali, tli 

 changes into he ; but in the spoken language, in some districts, the ku 



