THE EVOLUTION OF MALAY SPELLING. 93 



" in regard to a subject about which he is in such an infinitely 

 " better position to form a correct opinion. I was therefore at 

 " first inclined silently to accept as information his observations 

 " as to accent ; after further consideration, however, I found it 

 " preferable to come forward fearlessly with my doubts, and to 

 " expose myself if need be to a crushing- reproof, if this should 

 " be able to bring me, and perhaps others also, to a better view 

 '• on the subject. 



" I confess then that I have hitherto been under the convic- 

 " tion that in Malay the accent in derived as well as in root 

 ' k words fell as a rule upon the penultimate ; that on the addi- 

 " tion of a suffix the position of the accent changed as a rule 

 " from that which was originally the penultimate to that which 

 " was originally the last syllable : and that the pronunciation 

 " which according to Mr. van de Wall is the true and natural 

 " one, was quite peculiar to Europeans. It is so, I believe, in 

 " Javanese. It is true that there the accent does not come out so 

 " clearly as in Dutch, and in the Javanese grammar of T. Koorda 

 " (§ 87) the very existence of any accent in Javanese is denied : 

 " but what is there called 'a slower or more sleepy pronunciation 

 kb of the two last syllables of every polysyllabic word ' is more cor- 

 kw rectly interpreted, as it seems to me, as being a real accent on 

 " the penultimate, and a drawing out or longer holding on to 

 k ' the last syllable. Indeed, if one pronounces tulis, for instance, 

 u in the pure Dutch style with a clear accent on the penultimate, 

 " though this may not give the exact Javanese pronunciation, it 

 " is certainly much nearer to it than if one should say tufts, with 

 " an equally plain accent on the last syllable ; and similarly the 

 " pronunciation of the same word with the affix an, would, I 

 " believe, be better represented by tulissan or tulisan than by 

 " tulisan or tulisan. If the word is again increased by the addi- 

 " tion of another suffix, so that the original accented syllable is 

 " separated from the new suffix by one or more syllables, then 

 " besides the principal accent, which goes over to the last, the 

 " original accent again makes its appearance to some extent, as 

 " in tulisdne, rgdturuken. In a word, without digressing further, 

 " my proposition in the main is this : that in Javanese at any 

 " rate there actually is in the pronunciation of every word of 

 " two or more syllables a sort of stress, which can properly be 



