THE LANGUAGES OF THE INDIAN ARCHIPELAGO. 215 
each possessing many merits, and all illustrating the true 
principles on which our systems should be based, ought to sug¬ 
gest a scheme at once more simple* more uniform and more 
complete in the expression of vowel sounds than either. In the 
following attempt we do not presume that we have succeeded 
in this to any considerable extent, and we are aware of detects 
for which no remedy occurs to us. Our present object is 
fulfilled if we have obtained a means satisfactory to ourselves, 
of writing the languages of the Archipelago, and making our¬ 
selves intelligible to our readers. At the same time we hope 
its simplicity will recommend it to other writers in this 
Journal. 
In the three principal languages of the Archipelago, Malay, 
Javanese, Bugis, the most common vowel sounds are the Con¬ 
tinental a, i, u, o, and e, which, with Mr Crawfurd, we think 
ought to be expressed by the unaccented letters. The com¬ 
mon sounds of all the vowels admit of being contracted and 
lengthened, elevated and depressed. The short and long 
sounds occur in all the languages of the Archipelago and are 
generally invariable in each word, although often shifting for 
grammatical purposes, according to euphonic laws. The 
elevated and depressed tones are also found in all the lan¬ 
guages of the Archipelago, but they are not so essential as the 
others. They are however equally necessary to indicate the 
actual sounds of each language, and in comparative philology 
indispensible. The Archipelagic languages border on, and 
are allied to, the monosyllabic languages of the Hindu-Chinese 
countries, in which the tones are absolutely essential, since the 
meaning of a word alters with the tone * The only success- 
full attempt to express the tones of any of these languages of 
which we are aware, is that used by the Roman Catholic 
Missionaries in Cochin-china. It appears to be so good and 
so complete that we shall adopt it with some substitutions. 
The short and weak sound in every case we would express 
by the short prosodial mark. Mr Crawfurd dispenses with 
diacritical marks indicative of quantity or accent, upon 
the ground that the vowels are “long and short, or more 
correctly accented or unaccented according to their position 
in a word.” This however is not always the case even in Ma¬ 
lay, and we think one mark is necessary for each vowel to 
express the long and broad sounds often found in Malay, and 
which are characteristic of some dialects. In treating fully 
of any particular language exclusively, marks of accent and 
quantity need not in many cases be used, because all who are 
* The Mdayo —Poly sedan language! hate a few instances of tbla, 
