THE SINHALESE LANGUAGE. 



like this where all lexical analogies tend to establish a close affinity 

 to languages which are already ascertained to have sprung from a 

 Sanskrit source, I may indeed close the inquiry without at all 

 consulting Grammar. But, when with the evidence furnished by 

 the Dictionary we couple the testimony of History, and also find 

 historical facts confirmed by the analogies to which I have already 

 directed attention, it is impossible to resist the conclusion that the 

 Sinhalese is a legitimate descendant of the Sanskrit. 



Section Fourth. 

 Grammatical Relations. 



* The life and soul of a language, that which constitutes its sub- 

 stantial individuality, and distinguishes it from all others,' says 

 Professor Max Muller, 4 is its Grammar.' In accordance with this 

 undoubted belief, I puipose in this section, to examine the Gram- 

 matical forms of the Sinhalese with a view to ascertain whether they 

 have been imported from the South-Xudian, or from the Sanskrit 

 family of languages. In doing so I may as well intimate that I do 

 not intend to enter into an investigation of all grammatical forms, 

 but of such only as have been pointed out as possessing an intimate 

 relationship between the Dravidian and the North-Indian (in which 

 the Sinhalese has been included by some), and also a few of 

 such other forms as may throw light upon the inquiry in hand. 



The reader who has followed me through a variety of comparisons 

 of words, with overwhelming results in favor of the proposition with 

 which I have set out, must already be prepared to find the Gramma- 

 tical structure of our language to accord more intimately with the 

 Sanskrit than with the Dravidian. In this hope he will not indeed 

 be disappointed; but it is, perhaps, right to mention that the Sinha- 

 lese have also adopted some forms which bear some affinity to the 

 Dravidian. Founded upon a few coincidences between the Dravidian 

 and the North-Indian vernaculars, in which last I include the Sinha- 

 lese, it has been suggested that it would be more correct to represent 



